WHO/SDE/OEH 99.01 Original: English Distr. Limited EMF Risk Perception and Communication

Proceedings International Seminar on EMF Risk Perception and Communication Ottawa, Ontario, CANADA, 31 August – 1 September, 1998

Editors: M.H. Repacholi A.M. Muc

World Health Organization, Geneva, Switzerland 1999 Preface

ii Preface

EMF Risk Perception and Communication

iii Preface

iv Preface

EMF Risk Perception and Communication

Proceedings International Seminar on EMF Risk Perception and Communication Ottawa, Ontario, CANADA, 31 August – 1 September, 1998

Editors: M.H. Repacholi A.M. Muc

WHO/SDE/OEH 99.01

World Health Organization, Geneva, Switzerland

v Preface

U World Health Organization 1999

This document is not issued to the general public and all rights are reserved by the World Health Organization. The document may not be reviewed, abstracted, quoted, reproduced or translated, in part or in whole, without the prior written permission of WHO. No part of this document may be stored in a retrieval system or transmitted in any form or by any means - electronic, mechanical or other without the prior written permission of WHO.

The views expressed in documents by named authors are solely the responsibility of those authors.

vi Preface

Acknowledgement

Sincere thanks to the staff at the Radiation Protection Bureau of Health Canada; especially Drs. Elizabeth Nielsen and Art Thansandote. Julie Deschamps at the University of Ottawa and Alice McKinnon at Health Canada provided much needed assistance with the meeting organisation.

The support received from the World Health Organization, Health Canada, Industry Canada, Faculty of Medicine at the University of Ottawa, Cellular Telephone Industry Association, Canadian Wireless Telecommunications Association, B.C. Hydro, Ontario Hydro and Hydro Québec is gratefully acknowledged.

v Preface

Panel of Speakers and Chairpersons

M André Beauchamp Enviro-Sage Inc., Montréal, Québec, Canada

Dr William H Bailey Bailey Research Associates, New York, NY, USA

Dr Charon Chess Center for Environmental Communication, Cook College/Rutgers University, New Brunswick, NJ, USA

Mr Michael Dolan Assistant Director and General Counsel, EMF Advisory Group, ESAA Ltd., Melbourne, Victoria, Australia

Prof John D Graham Harvard Center for Risk Analysis, Harvard School of Public Health, Boston, MA, USA

Dr Philip Gray Programmgruppe Mensch, Umwelt, Technik (MUT), Forsuchungszentrum Jülich GmbH, Jülich, Germany

Mrs Ruth Greey, Ontario Hydro, Toronto, Ontario, Canada

Mr Gerry Kruk Gerry Kruk & Associates Communications Ltd., Calgary, Alberta, Canada

Ms Judy Larkin REGESTER LARKIN, London, UK

Prof William Leiss Environmental Policy Unit, School of Policy Studies, Queens University, Kingston, Ontario, Canada

Dr Tom McManus Department of Public Enterprise, Dublin, Ireland

vi Preface

Dr Elizabeth Nielsen Radiation Protection Bureau, Health Canada, Ottawa, Ontario, Canada

Dr Chris Portier Chief, Computational Biology and Risk Analysis, NIEHS, Research Triangle Park, NC, USA

Dr Michael H Repacholi, Department of Protection of the Human Environment, World Health Organization, Geneva, Switzerland

Dr Dan Wartenberg Environmental and Occupational Health Sciences Institute, Pisscataway, NJ, USA

Dr Peter M Wiedemann Gruppe MUT, Forsuchungszentrum Jülich, Jülich, Germany

vii Preface

Foreword

This international seminar entitled “EMF risk perception and communication” was held at the University of Ottawa, Ontario, Canada from 31 August to 1 September 1998. The seminar was sponsored jointly by the World Health Organization (WHO), Health Canada, Industry Canada, the Faculty of Medicine at the University of Ottawa, the Cellular Telephone Industry Association, the Canadian Wireless Telecommunications Association, BC Hydro, Hydro Quebec and Ontario Hydro.

Possible health effects of exposure to electromagnetic fields (EMF) have led to concerns among the general public and workers that appear to go well beyond those that are attributed to well-established risks. It is necessary to understand why this occurs and to deal with it through an effective communications programme.

People have the right to access reliable, credible and accurate information about any health risks from EMF exposure. However, recent history has shown that, communication among scientists, governments, industry and the public, has often been ineffective. There continues to be a divergence of views because of this failure to communicate effectively.

A hazard is a set of circumstances or a situation that could harm a person's health, while risk is often defined by scientists as the likelihood (or probability) that harm will occur from a particular hazard. Living is a risky business. Every activity has an associated risk. When you awake you may fall down the stairs. While cooking breakfast you may be electrocuted. Driving to work can result in a car accident. People will generally take risks if they perceive that there is some advantage or benefit. Normally the benefits should outweigh the risks by a significant margin. However, there is no such thing as a zero risk.

How people perceive risks can depend on their age, sex and cultural background. Many young people are happy to go sky diving. Older people are generally not attracted to this activity since they perceive it as too dangerous. In addition to balancing costs and benefits, when determining the acceptability of a risk, people may compare it with something usually acceptable in normal life. This can lead to risk being perceived at various levels: negligible, acceptable, tolerable, or unacceptable.

viii Preface

The nature of the risk can also lead to different perceptions. Surveys have found that the following pairs of characteristics of a situation generally affect risk perception. The first member of the pair tends to increase and the second, to decrease the magnitude of the perceived risk.

Involuntary vs. Voluntary exposure. This is an important factor in some persons’ perception of the risk from EMF exposures, especially from fixed RF antennae or ELF transmission lines. Mobile telephone users’ generally have a lower perception of risk from the much more intense RF fields emitted by their voluntarily-chosen handsets, while non-users have a higher perception of risk from the relatively low RF levels emitted from mobile telephone base stations.

Lack of personal control vs. Feeling of control over a situation. People do not normally have any say about installation of EMF facilities, such as power lines or mobile telephone base stations. This can then raise their concerns.

Familiar vs. Unfamiliar or exotic. Familiarity with the situation, or a feeling of understanding of the technology, helps reduce risk perception. Perception of risk is increased in a new, unfamiliar, or hard-to-comprehend situation or technology, such as EMF technology. This perception can be significantly increased if there is a feeling that science does not fully understand the situation, as with the current uncertainty in the data concerning possible health effects from low-level EMF exposure.

Dreaded vs. Not dreaded. Some disease outcomes, such as cancer, severe and lingering pain, disability, or genetic threats to future generations, are more highly feared. Thus possible small increases in cancer, especially in their children, from EMF exposure receive significant attention.

Unfairness vs. Fairness. An unfair risk is different from one that is involuntary, though that factor may also be considered unfair. Unfairness may come from an uneven distribution of risk across the population. If people are exposed to RF fields from mobile telephone base stations, but do not have a mobile telephone or they are exposed to the electric and magnetic fields from a high voltage transmission line which only provides power to people further along the line, then people consider these situations unfair and are less likely to accept any associated risk.

ix Preface

Taking the example of mobile telephone base stations, especially for people not owning a mobile telephone, the EMF emissions have the following factors:

• They are an involuntary exposure that people feel they do not need.

• There is a lack of control over their installation (especially its location).

• The technology is unfamiliar to most people.

• Scientists do not have enough information to be precise about health risks, and there is the possibility that it could cause a dreaded disease such as cancer.

• Installation of these base stations, which expose everyone to EMF, only benefits people who use mobile telephones, and so it is unfair.

Gone are the days when scientists could assess a risky situation and the public would accept this analysis without question. Unfortunately there has been a decline in respect of expert opinion. Research has shown that effective communications depends on the establishment of trust and credibility of the expert (s) and the sources of information.

Trust and credibility are key to the process of risk communication. They are hard to gain and maintain, but are easily lost and, once lost, much harder to regain. While acting in a trustworthy and credible manner is necessary, that alone will not guarantee that trust will be gained. Perceptions, including perceptions of past action, are also important. Other factors include empathy and caring, competence and expertise, honesty and openness, and dedication and commitment. In a democratic society people feel they have a right to know what is proposed and might affect them directly. They want to have some control and be part of the decision-making process.

The International EMF Project is preparing reports that will provide strategies for understanding the concerns of all parties in the EMF debate and the elements of an effective risk communications programme. This international seminar is the second on risk perception, communication and management to be held within this Project and will lead to the following outputs:

• This proceedings of all presentations;

x Preface

• An Environmental Health Monograph providing the basic science of risk perception, communication and management; and

• A user-friendly handbook that addresses this issue with respect to EMF.

Copies of these documents when printed will be available from:

International EMF Project Department of Protection of the Human Environment World Health Organization CH-1211 Geneva 27, Switzerland Fax: + 41 22 791 4123

The Editors February 1999.

xi

Preface

Contents

Preface

Acknowledgement V

Panel of Speakers and Chairpersons VI

Foreword VIII

Keynote Presentation

Making Sense of Risk 1 Dr. John D. Graham

Presentations

EMF Concerns and WHO’s International EMF Project 29 Dr. Michael H. Repacholi

Improving EMF Risk Communication and Management: 51 The Need for Analysis and Deliberation Dr. Philip Gray

Dealing with EMF Risk Perceptions: Themes, Challenges 69 and Potential Remedies Dr. Peter Wiedemann

Risk Communications and Managing EMF Controversies 95 Mr. Gerry Kruk

Evaluating Response Options 119 Ms. Judy Larkin

A Public Controversy in Canada Over Health Risks 133 Associated with Radiofrequency Fields Prof. William Leiss

xiii Preface

EMF and Public Consultation 157 Mr. André Beauchamp

Principles of Risk Assessment: 173 Application to Current Issues Dr. William Bailey

How Dangerous Is It Really? 191 Some Approaches for Inferring Risk Dr. Dan Wartenberg

NIEHS EMF Risk Assessment Program 205 Dr. Christopher Portier

Public Participation: Practice Based on Research 223 Dr. Charon Chess

Elements of a Prudent Avoidance Policy 237 Dr. Tom McManus

Prudent Avoidance: 253 Does It Have Application in EMF Risk Management? Mr. Michael Dolan

The Politics of EMFs - An Activist's Viewpoint 279 Mr. Richard W. Woodley

Summary 281 Dr. Tom McManus

xiv Preface

xv Preface

xvi EMF Risk Perception and Communication

Keynote Presentation

MAKING SENSE OF RISK: An Agenda for Congress

John D. Graham, Ph.D. Director, Harvard Center for Risk Analysis Department of Health Policy and Management School of Public Health, Harvard University 718 Huntington Avenue, Boston, Massachusetts, 02115, USA Tel: (617)432-4343, Fax: (617)432-0190

from: RISKS, COSTS, AND LIVES SAVED: Getting Better Results from Regulation Robert W. Hahn, ed., Oxford University Press, 1996. (Reprinted with permission)

OVERVIEW

The American people are suffering from what can be called "a syndrome of paranoia and neglect" about potential dangers to their health, safety, and the environment. This leads to a paradox that is becoming increasingly recognised. Large amounts of resources are devoted to slight or speculative dangers while substantial and well-documented dangers remain unaddressed.

Congress can take a modest step toward curing that syndrome by embracing a risk-analysis approach to public decision making. Omnibus legislation covering federal agencies should include:

• requirements for responsible quantitative risk assessment before making protective decisions

• periodic risk rankings for setting priorities on the basis of science and citizen preferences

• public reporting of estimated risks, costs, and benefits of new rulemakings

1 EMF Risk Perception and Communication

• external peer review of analyses to enhance quality and credibility

• a presumptive burden that costs are reasonably related to benefits (without insisting on full-scale quantification, monetisation, and positive net benefits)

• the opportunity for affected citizens to seek judicial review in the event that agencies do not use the risk-analysis framework.

Such omnibus legislation, applying to the entire regulatory process, would greatly improve outcomes, but it must be crafted carefully. This chapter outlines components of such legislation.

To avoid "paralysis of analysis," Congress should authorise agencies to tailor the intensity of analysis to the importance and complexity of the specific problems. Congress should provide adequate budgetary and technical resources for agencies to discharge their analytical functions.

Statement of the Problem

Each day Americans are confronted with new information about potential dangers to their health and safety: childhood cancer from exposure to electric and magnetic fields, male infertility from exposure to chlorinated chemicals, brain cancer from using cellular telephones, subtle neurologic effects in children from ingesting small amounts of lead in house dust, premature death from inhaling fine particles in urban air, heart disease from ingesting margarine and other sources of transfatty acid, and non-Hodgkin's lymphoma from exposure to phenoxy herbicides on the farm.

Although most objective indicators suggest that America's burden of mortality and morbidity risk is steadily declining (Department of Health and Human Services annual), citizens are confused about the growing number of hazards in daily life reported in the media that appear to have some degree of scientific support (Singer and Endreny 1993). Misperception of risk may be widespread. For example, a majority of Americans perceive that "things in the environment" are at least as important as "personal habits" in causing sickness and poor health (Dunlap 1991). Yet the best available scientific data indicate that personal habits are a much more important cause of poor health than environmental agents (McGinnis and Foege 1993).

2 EMF Risk Perception and Communication

The public's general reaction to health, safety, and environmental dangers may best be described as a syndrome of paranoia and neglect. We are paranoid in the sense that we devote large amounts of resources and attention to alleged dangers that are speculative (at best) and probably small (or even nonexistent). Examples of “overblown" hazards include soil and groundwater contamination at many abandoned hazardous waste sites (Environmental Protection Agency 1990; National Research Council 1991a), the pesticide residues on fruits and vegetables purchased in grocery stores (Ames 1992; Archibald and Winter 1990), the benzene in the ambient air of urban and rural communities (Graham 1993), and the residual chlo- roform found in drinking water after disinfection of water supplies through chlorination (Larson, Wolf, and Butterworth 1994). None of those hazards constitutes a major public health problem, but the media and government agencies treat them as if they are.

Accompanying that paranoia is a disturbing degree of tolerance of well- documented and substantial dangers to public health and environmental quality. Examples of "neglected" hazards include violence in families and communities (National Committee for Injury Prevention and Control 1989), deteriorating lead paint in older homes (Florini and Silbergeld 1993), inadequate use of basic preven- tive health services such as childhood immunisations, influenza vaccinations, and breast cancer screening, and hazardous lifestyles characterised by smoking, abuse of alcohol, high-fat diets, lack of physical exercise, and failure to use basic safety devices such as smoke detectors and lap/shoulder belts in cars (U.S. Department of Health and Human Services 1990). These are all major public health problems that receive less than their fair share of attention in media stories and public policy.

Even within a single domain of public policy, such as environmental health, the syndrome of paranoia and neglect is evident. In 1990 Congress passed 1,200 pages of amendments to the Clean Air Act Amendments of 1970 that are estimated to cost the nation an additional $30 billion per year (Portney 1990). The beneficial consequences of that intensified effort to reduce outdoor air pollution are considerable and include a variety of ecological, aesthetic, and human health benefits. Nonetheless, there is a growing scientific consensus that exposure to indoor air pollution is a more serious human health problem than exposure to outdoor air pollution (Samet and Spengler 1991; National Research Council 1991b). Yet no powerful political demand or sustained advocacy effort has aimed at new legislation to assess and improve indoor air quality.

3 EMF Risk Perception and Communication

The syndrome is also revealed by simple comparisons of the policy debates about health care reform and environmental policy. Consider, for example, the relative cost-effectiveness of reducing cancer through pollution prevention versus early detection and treatment of tumours.

Mrs. Clinton and her task force proposed to Congress a basic health security package that included coverage of preventive screening every two to three years for breast and cervical cancer. The estimated cost of those measures is as low as $1,000 to $10,000 per year of life saved. More frequent screening intervals, say once a year, would have offered slightly more protection, but at an incremental cost in excess of $100,000 per life-year saved (Eddy 1990). Mrs. Clinton's proposal was therefore a form of "implicit rationing" of insurance coverage.

At the U.S. Environmental Protection Agency, the "shadow prices" for an extra life-year are far more extravagant than what Mrs. Clinton proposed. For example, the EPA recently justified a series of new regulations aimed at reducing benzene emissions from various industrial sources on the grounds of cancer prevention (with no stated expectation of accompanying benefits from diminished noncancer health effects or ecological protection). The cost of those benzene rules, using EPAs figures, ranges from $200,000 to $50,000,000 per year of life saved (EPA 1989).

In defence of pollution prevention, citizens would certainly prefer to prevent a tumour from forming rather than simply treat it early, since a family experiences considerable emotional trauma when a tumour is detected. If we knew that particular pollutant exposures were increasing the risk of breast cancer, the case for primary prevention would certainly merit priority attention. Yet one wonders whether citizens would support up to fifty fold differences in the fiscal priorities that the government is now placing on life-years saved by different cancer prevention policies (Mendeloff and Kaplan 1989).

Winds of Change

The syndrome about risk will not be easy to change because to do so will require us to re-examine our current mental models of risk. Change may also require challenges to various power structures in Washington, D.C., and elsewhere that have capitalised and prospered from the syndrome. Fortunately, a small but growing coalition of reformers is calling attention to the syndrome about risk.

4 EMF Risk Perception and Communication

Scientists are questioning the logic behind traditional policies of risk assessment and management. For example, biochemist Bruce Ames of the University of California at Berkeley has raised serious questions about whether findings of carcinogenicity in high-dose animal tests are valid indicators of dangers to people from low-level exposures to chemicals in the environment (Ames, Profet, and Gold 1990). Less vocal yet influential concerns have also been expressed by committees of the National Academy of Sciences and selected leaders of the American Association for the Advancement of Science (National Research Council 1992, 1994; Abelson 1993).

In the all-important mass media, Keith Schneider of the New York Times, Boyce Rensberger of the Washington Post, David Shaw of the Los Angeles Times, and John Stoussel of ABC News have begun to expose the syndrome in highly visible stories and documentaries. Those efforts are stimulating reporters and the public to ask harder questions about which dangers receive media coverage and which ones are overblown and neglected.

Governors and mayors, citing the "unfunded mandates" emanating from Washington, D.C., are frustrated about their lack of freedom to allocate resources to the dangers that are most evident and prevalent in their states and communities. In Columbus, Ohio, for example, environmental mandates from the EPA are forcing cuts in basic public health services typically provided by the state's health depart- ment (Environmental Law Review Committee 1991). The mayor of Anchorage, Alaska, has engaged in a concerted effort to expose the distorted view of health problems that emerges from the federal government's view of risk (Municipality of Anchorage 1992). Since the probability of major tax increases to expand governmental services is low, the sobering reality is that resources for protection against dangers are being rationed - even during good economic times!

Business firms, both large and small, are questioning whether the speculative risks of industrial activities should be regulated to vanishingly small levels without regard to the economic costs of such regulations. The growing costs of the tort liability system, which also reflect public perceptions of risk as much as science, are also a significant cost to important segments of industry (Foster, Bemstein, and Huber 1994). Since American firms operate increasingly in a competitive global economy, each addition to the cost structure of American firms - including those generated by a "better safe than sorry" approach to risk regulation and liability - is coming under increasing scrutiny by government as well as industry (Business Roundtable 1994).

5 EMF Risk Perception and Communication

In the deliberations of the 102d Congress, one can see the first clear indications that governmental mismanagement of risk has emerged as a national political issue. The indications began with a straightforward effort by Senator Bennett Johnston (D-Louisiana) to require risk assessment, risk comparison, and benefit-cost analysis as an amendment to a bill calling for the elevation of the EPA to cabinet status in the executive branch. That amendment passed on the floor of the Senate by a vote of 95 to 3.

In the House of Representatives, the Clinton administration and the Democratic leadership made a sustained effort to avoid a Johnston style risk amendment on the theory that a "clean bill" would be less likely to offend certain advocacy groups or legitimise a string of more controversial amendments. When the House Democratic leadership at the urging of Vice President AI Gore and EPA Administrator Carol Browner, sought to impose a procedural rule blocking such amendments, they were stunned to lose the floor vote by a considerable margin. A loose yet potent bipartisan coalition of representatives led by Gary Condit (D- California), Karen Thurman (D-Florida), Pete Gerer (D-Texas), John Mica (R- Florida), and Richard Zimmer (R-New Jersey) expressed their determination to pass significant risk-related legislation.

Since those two floor votes occurred, proper treatment of "risk" has become a major legislative issue. Numerous bills calling for more or better risk analysis were introduced in both the House and Senate. Risk-related amendments were prominent in debates over reauthorization of Superfund, the Safe Drinking Water Act, the Clean Water Act, the U.S. Department of Agriculture, and even an environmental technologies bill. Republicans and Democrats are beginning to compete to take credit for championing the emerging "risk issue," and new risk- protection legislation is unlikely to pass until Congress has the opportunity to consider a new risk-based approach to managing public health and environmental dangers.

What Kind of Change?

While there is a growing political consensus that something needs to be done, the hard thinking has only begun about what Congress can do to help cure the country's syndrome of paranoia and neglect. An implicit assumption of the reform movement is that the federal government has failed to conduct risk analyses or make proper use of risk analysis in public policies aimed at protecting human health, safety and environmental quality.

6 EMF Risk Perception and Communication

The phrase risk analysis, following the interdisciplinary interests of the Society for Risk Analysis, refers broadly to the related analytical tools of risk assessment, risk characterisation, risk comparison, risk ranking, risk-based priorities, risk management, risk-benefit analysis, benefit-cost analysis, cost-effectiveness analysis, decision analysis, and risk communication.

Although there are important and subtle differences in these tools (Merkhofer 1987), the term risk analysis often refers to the collection of them.

But risk analyses of various sorts are already used widely (if not uniformly and consistently) by agencies such as the Consumer Product Safety Commission, the Department of Energy, the Department of Transportation, the Environmental Protection Agency, the Nuclear Regulatory Commission, and the Occupational Safety and Health Administration. In fact, the Clinton administration's 1993 executive order on regulatory planning requires risk analyses to be submitted in support of all significant rulemaking initiatives. It is important therefore to be clear about what specific legislative steps are needed to attack the syndrome of paranoia and neglect.

Responsible Risk Assessments

When potential dangers are brought to the attention of federal agencies, agencies need to assess those dangers in a responsible manner. Congress can work to inculcate a strong sense of responsibility by requiring agencies to follow several basic principles of sound risk assessment practice.

First, Congress should compel agencies to make use of the best available scientific information. While that expectation may seem like nothing more than a plea for "apple pie and motherhood," agencies occasionally neglect or reject high- quality scientific information. They do so because they may prefer the apparent consistency provided by use of "default" assumptions and models and because bureaucracies have difficulty innovating without a clear statutory mandate. The EPA, for example, is only beginning to use innovative biological studies that suggest that low doses of unleaded gasoline vapours, chloroform, and formaldehyde pose less risk to people than previously thought (Graham 1991a). When new scientific studies suggest that a hazard is more dangerous than previously thought (for example, dioxin and fine particles), agencies also tend to be slow to respond to new information. Some degree of skepticism about new discoveries is appropriate, but

7 EMF Risk Perception and Communication the government's current risk-assessment process can hardly be accused of being overly responsive to scientific developments.

Second, when scientific knowledge about risk is imperfect or deficient, Congress should require agencies to employ probabilistic methods of uncertainty analysis. Agencies should report a range of risk estimates, accompanied by each estimate's likelihood of being correct, to decision makers in publicly available documents (Morgan and Henrion 1991). Single-point estimates, such as plausible upper bounds or worst-case scenarios, should generally be accompanied by lower, bound (optimistic) and realistic (or likely) estimates of risk (Graham 1991b), unless a worst-case figure is being used strictly for "screening purposes" (where the objective is to rule out exposures or hazards that are too tiny to worry about). Agencies should follow an iterative procedure that calls for more detailed uncertainty analysis as the importance of the decision increases (National Research Council 1994). Qualified experts in formal uncertainty analysis should be recruited and trained to participate in the preparation of risk assessment reports (Cooke 1991). Some agency scientists and advocates may perceive that only worst-case estimates of risk will induce regulators to provide the maximum degree of public health protection, but Congress needs to insist that value judgements about the proper margins of safety be made by accountable regulators and not be buried in the choice of particular assumptions or models used to compute risk (Federal Focus Inc. 1991; Zeckhauser and Viscusi 1990).

Third, when the same hazard poses more danger to some citizens than others, Congress should insist that agencies report that information through distributional methods of variability analysis. For example, some citizens are more sensitive to environmental agents than others for genetic or lifestyle reasons (Brain, Beck, Warren, and Shaikh 1988). Some citizens are also exposed more to hazards than others (for example, those who live directly downwind from a factory). Agencies should present to decision makers a public document with information about the number of citizens exposed to various levels of risk. Since low-income and minority citizens often incur a disproportionate share of public health and environmental risks, agencies should make a special effort to investigate those citizens' degree of exposure and susceptibility to hazards (Zimmerman 1993). Without that kind of information, risk managers are in a poor position to incorporate equity and justice considerations into their decisions1.

1 Interestingly, several years ago the federal government embraced a form of „variability analysis“ of economic costs when it required that regulatory impacts on small businesses be highlighted in rulemaking notices.

8 EMF Risk Perception and Communication

Fourth, to nurture the public's sense of perspective about risk, Congress should require agencies to make thoughtful use of risk comparisons (Wilson 1979). Analogies are a powerful communication and learning tool if they are crafted with forethought. For example, when agencies Congress needs to insist report that the extra cancer risk from eating that value judgements about pesticide residues on food is one in a proper margins of safety be made million lifetimes, they should state how by accountable regulators and large this risk is compared with other risks not be buried in the choice of in daily life. A baby born today, at current particular assumptions or models mortality rates, incurs a risk of four in a used to compute risk. million of being struck and killed on the ground by a crashing airplane (Goldstein, Demak, Northridge, and Wartenberg 1992). Although the airplane hazard is far better documented than the estimated risk of ingesting pesticide residues, such a comparison can help citizens and journalists develop an intuition about relative magnitudes. Since the purpose of those risk comparisons is educational and is not an explicit part of a formal priority-setting process, it is neither necessary nor appropriate for an agency to restrict the comparisons to hazards that happen to fall within its jurisdiction. As long as the comparison does not force or constrain any value judgement about acceptability, it is not crucial for the agency to ensure that the risks are comparable in terms of other dimensions such as controllability or preventability (Roth, Morgan, Fischhoff, Lave, and Bostrom 1990). If an agency is making a value judgement about acceptability, then it should also consider dimensions of risk other than numerical magnitude (Fischhoff, Lichtenstein, Slovic, Derby, and Keeney 1981).

Finally, Congress should require agencies to assess a broad range of potential human health and environmental effects. Historically, agencies have tended to focus on mortality effects (especially from cancer). While premature death is obviously important, citizens are concerned about a much broader range of effects on functional status and quality of life. As the EPAs proposed reassessment of dioxin has indicated, the levels of exposure associated with negligible cancer risk are not necessarily low enough to eliminate concern about potential threats to the immune and reproductive systems (EPA 1994a). Providing protection for human health also does not always ensure protection of nonhuman life forms and ecosystems (Hegner 1994). In the field of traumatic risks, federal agencies are also beginning to recognise the long-term functional consequences of nonfatal head and brain injuries (National Highway Traffic Safety Administration 1994).

9 EMF Risk Perception and Communication

As Congress becomes more serious about promoting sound risk assessment practice, agencies will need adequate budgetary and technical resources to get the job done. Under some conditions, it may make sense to move the assessment function outside the mission-oriented regulatory agencies to achieve a greater degree of objectivity and credibility. Again, such organizations will have extra start-up costs, will require adequate budgetary and technical resources, and will complicate the challenge of co-ordinating risk assessment and management (National Research Council 1993). For assessment units to provide attractive career paths for young scientists, they must also have access to exciting intramural and extramural research and training opportunities - an aspect of risk analysis that currently exists nowhere in the federal government. Hence, the emerging coalition for reform of risk regulation needs to be sensitive to the analyst's need for various kinds of resources. Without such commitments, we cannot expect significant and sustained improvements in analytical practice.

In the 102d Congress, improved risk assessment was the subject of important bills introduced by Representatives Carlos Moorehead (R-California) and Herbert Klein (D-New Jersey). Then, after the 1994 election, the momentum for reform increased in the 103d Congress, and by late summer of 1995 very significant legislation had passed the House and was under consideration in the Senate. The House-passed bill (H.R. 1022), with a few revisions, will provide the public with more protection against genuine risks at less total cost than under present law. That is the case despite often-emotional attacks on the bill by journalists, environmentalists, and others.

Although the House bill has flaws - for example, its obtrusive standard of court review - it is based on sound scientific principles and provides regulators with the necessary flexibility to make tough calls about specific hazards.

Risk-Based Priority Setting

Although the federal government undertakes numerous risk analyses each year, few of them address the "big picture" questions about how resources are allocated among various dangers. For example, why are we spending billions of dollars cleaning up lead in soil at industrial sites, where the probability of childhood exposure is low, when we spend few resources to protect urban children against the neurotoxic effects of ingesting deteriorating lead paint in old homes? In the absence of risk rankings, it is difficult for Congress, regulatory agencies, and the public to

10 EMF Risk Perception and Communication gain a sense of perspective about the relative importance of each new danger reported in the mass media.

To counteract the "risk-of-the-month" syndrome, Congress should require the executive branch to periodically rank hazards according to their seriousness and the available opportunities for cost-effective reduction. If such rankings were publicly available, reporters could frame questions and write stories about how newly alleged hazards might rank in seriousness relative to hazards that have already been ranked. For example, which is a more serious threat: contracting cancer from inhaling environmental tobacco smoke or contracting cancer from exposure to radon in the basements of homes? Each agency should be responsible for ranking hazards that fall within their jurisdiction, while the Office of Science and Technology Policy should work with an interagency committee to produce a general ranking of hazards (Carnegie Commission on Science, Technology, and Government 1993).

Since risk-ranking exercises are technically demanding and require delicate value judgements, agencies should undertake them with care and in accordance with well-considered guidelines. Congress should authorise the OSTP and the Office of Management and Budget to develop jointly guidelines governing the proper conduct of risk-ranking exercises. Agencies should also issue an "annual report on risk" indicating how their allocation of resources reflects their ranking of hazards, including what they accomplished in the previous year.

The OMB should also use the agency rankings as a contribution to continuing discussions about agency budgets - a modest form of "risk-based budgeting" - including potential limitations on the "offbudget" or unfunded expenditures made by states and localities and the private sector. Trends in staffing and budgets for risk-protection agencies have never been compared with scientific information on agency performance in problem solving. The years from 1984 to 1994 were good for the EPA but fairly lean for most other risk-protection agencies (Warren 1994). Yet there is certainly no hard evidence that the EPAs performance in cost-effective risk reduction has been superior to the performance of other risk- protection agencies.

In recent years the EPA and many state environmental agencies have acquired significant experience in risk-ranking exercises (EPA 1993b). On the basis of that experience, Congress should insist that agencies develop guidelines on several critical issues.

11 EMF Risk Perception and Communication

First, there is at least some degree of scientific uncertainty about the seriousness of every hazard. In some cases, such as the catastrophic consequences of rapid global warming from greenhouse gases, the uncertainties are so large that analysts wield enormous discretion in the assessment and ranking process. If "worst-case" estimates of risk are used to characterise hazard A while realistic estimates of risk are used to characterise hazard B, the ranking of risks will be biased toward placing hazard A at the top of the ranking. While there are many valid approaches to resolving that issue (for example, reporting ranges or probability distributions that reflect uncertainty, or on occasion simply refusing to rank poorly understood hazards), Congress should compel the OSTP and the OMB to develop guidelines about treating uncertainty consistently.

Second, informed lay persons as well as experts must explicitly acknowledge and consider the value judgements in risk ranking. Cognitive psychologists have demonstrated that lay people often harbour genuine concerns about risks that are not captured by standard technical estimates of risks (Fischhoff et al. 1978). For example, if one hazard creates 100 "involuntary" cancers, while another hazard creates 1,000 "voluntary" cancers, which hazard should be assigned the higher ranking? Public participation by interest groups is not always the best approach, since there may be a clear incentive to “game" the value-judgement process in a way that favours industry, environmentalists, or consumers. A better approach may be to elicit critical information on value judgements (for example, the relative importance of an equal number of cancer, trauma, and neurologic effects) from a representative citizen panel objectively selected by a panel of social and physical scientists (Fischhoff 1994).

Considerable scientific progress has been made in formally eliciting thoughtful value judgements. Those techniques are already beginning to influence the practice of medicine through the physician-patient relationship (Drummond, Stoddart, and Torrance 1987). When postmenopausal women decide whether to undergo long-term estrogen therapy, they are making a value judgement about the relative importance of cancer risks, which therapy increases, versus hip fracture and heart disease risks, which therapy reduces. Congress should require the OSTP and the OMB to develop guidelines on the development and application of those forms of science-based "public participation" in risk ranking.

Finally, since optimal risk reduction is the ultimate goal of risk policy, the risk-ranking process should include a component that entails considering the relative cost-effectiveness of alternative measures to reduce risk. The most serious dangers

12 EMF Risk Perception and Communication should not always be ranked the highest since the available risk-reduction strategies may be fairly ineffective or costly, although research into new interventions may merit high priority. Risks of only a moderate degree of seriousness may be ranked high if it is feasible to eliminate them at a low cost to society. Congress should require the OMB and the OSTP to develop guidelines on how to incorporate considerations of policy effectiveness and cost into the process of setting risk-based priorities (Graham and Hammitt forthcoming).

Report Risks, Benefits, and Costs

When policy makers work to reduce risks, they should routinely quantify the target risks and consider what benefits and costs are anticipated to result from their favoured policies. Legislators in particular need to recognise the need to assess risks, benefits, and costs before passing new laws. Many states have ignored that principle in laudable efforts to promote "pollution prevention" and "toxics use reduction" (Laden and Gray 1993). Unfortunately, those approaches offer no analytic framework for deciding whether it makes sense to spend scarce resources substituting chemical A for chemical B. The Clinton administration's proposed study and phase-out of chlorine suffers from a similar problem: there is no requirement for a comparative economic evaluation of chlorine and its likely substitutes in various uses!

The administration's chlorine initiative does call for a comparison of the relative safety of chlorine and its substitutes. Congress must embrace the principle of considering the competing risks of regulation as well as the target risks in broad- based legislative language (Graham and Wiener 1995). Congress should do so either by clarifying that the risks induced by regulation are "costs" or by clarifying that the risk assessment report should include substitution risks as well as target risks.

Congress needs to mandate consideration of risk trade-offs because agencies sometimes have incentives to downplay risks that are being induced by their policies. In the 1980s, for example, the National Highway Traffic Safety Administration refused to acknowledge that automobile fuel economy rules were decreasing occupant safety by encouraging vehicle manufacturers to build smaller and lighter cars (Graham 1992). NHTSA is beginning to acknowledge that competing risk but only reluctantly, following a federal court order (Competitive Enterprise Institute v. National Highway Traffic Safety Institute, 956 F.2d 321 (D.C. Cir. 1992)).

13 EMF Risk Perception and Communication

President Clinton's 1993 executive order on regulatory planning does require federal agencies to analyse competing risks, costs, and benefits, but only if the proposed rule is a "significant one" - for example, incurring national costs in excess of $100 million. When the statutory criteria governing an agency’s decision do not permit the agency to weigh the benefits and costs of alternatives, the agency may be inclined not to invest the time and resources in a risk analysis - a decision that the Clinton administration has been inclined to tolerate.

It is useful to examine in some detail an example of the problems and flaws in current practice that the suggested omnibus legislation should correct. A good example - it is by no means the only one, but it illustrates the point - is the EPAs implementation of Title III of the Clean Air Act Amendments of 1990. I briefly contrast that with the more reasonable procedures of the National Highway Traffic Safety Administration.

In Title 111 Congress required the EPA to promulgate technology based standards for numerous industrial source categories that emit so-called hazardous air pollutants. On the basis of a survey of the rules proposed and finalised to date under Title 111, it appears that the EPA is reporting costs but making less effort to report benefits.

In one of the first rules finalised under Title Ill, the EPA required technological controls of 110 different hazardous organic pollutants in the chemical manufacturing industry. The EPA reported that the rule would induce up-front capital expenditures of $450 million and net annualised costs of $230 million per year. The official rulemaking notice includes no estimate of the rule's benefits, except that the mandated controls would reduce the quantity of hazardous air pollu- tants by 510,000 tons per year and the quantity of other volatile organic compounds by 490,000 tons per year. The EPA provides no information about the target risks or the number of adverse health effects prevented. The EPA does refer to a regulatory analysis in the public docket prepared at the OMB's request, but the cover page of that document indicates that insufficient time and data were available to produce "scientifically supportable" estimates of benefit (EPA 1994c).

In the case of a similar proposed rule under Title Ill covering 190 petroleum refineries in the United States, the EPA reported initial capital costs of $207 million and net annualised costs of $110 million per year. The estimated reduction in the quantity of pollutants was 54,000 megagrams per year of hazardous air pollutants and 300,000 megagrams per year of other volatile organic compounds. An accom-

14 EMF Risk Perception and Communication panying regulatory analysis in the public docket projects that the rule will avert less than one case of cancer per year.2 The EPA does project $148.3 million per year in health benefits from the reductions in emissions of volatile organic compounds, assuming that each megagram of volatile organic compound is assigned a dollar value. The agency acknowledges that the uncertainty in this critical "shadow price" ranges from $25 to $1,574 per megagram. That wide range was based on a secondary source (the Office of Technology Assessment), with no independent peer review by health scientists or health economists on the agency's Science Advisory Board. Even when an agency develops a rule through a "regulatory negotiation" involving affected interests, the agency does not necessarily consider the rule from the perspective of social risks, costs, and benefits. In 1993 the EPA finalised a negotiated rule aimed at reducing hazardous air pollutants at thirty coke production plants through two phases of technological controls: the maximum achievable control technology and the lowest achievable emission rate. The EPA reports that the up-front capital cost of the maximum achievable control technology for existing plants will be $66 million, with net annual costs of $25 million. The standards for the lowest achievable emission rate may require $510 million in capital costs and net annualised costs of $84 million. The final rule reports no estimate of risks from hazardous air pollutants or benefits from their reduction, although a 1992 study commissioned by the EPAs Office of Policy Analysis estimated that the maximum achievable control technology and the lowest achievable emission rate rules together would avert less than one case of cancer per year (Considine, Davis, and Marakovits 1992). The EPA concluded that this rule does not require OMB review, in part because it does not cost more than $100 million per year (EPA 1993a).

Most recently, the EPA issued a proposed rule under Title III covering the aerospace industry that would impose $503 million to $714 million in up-front capital costs and net annualised costs of $16.7 million. The agency reported no risk estimates or health benefits except an estimate that the proposed rule would prevent 208,000 tons per year of hazardous air pollutants. The EPA stated that since annual costs were less than $100 million and since Congress did not require benefit-cost analysis, the agency would not report a regulatory analysis for OMB review (EPA 1994b).

Other risk-protection agencies do not share the EPAs proclivity to impose expensive regulations on the economy without conducting any serious analysis of their benefits. NHTSA, for example, reports a rough estimate of costs and benefits

2 Other categories of possible health benefits related to the reduction of hazardous air pollutants were noted but not quantified.

15 EMF Risk Perception and Communication for every rulemaking, regardless of whether its cost exceeds $100 million3. That practice reflects more the agency's analytical tradition than any specific statutory requirement to perform benefit-cost analysis of all rules. To prevent analytical burdens from creating excessive delays in rulemaking, NHTSA tailors the intensity and precision of the regulatory analysis to the importance of the rule. Thus, less important rules receive less detailed analyses.

The decision not to invest in benefits analysis will ultimately make it difficult for the EPA to present a persuasive case to Congress and the public that their rulemakings on hazardous air pollutants have been worthwhile. Interestingly, an explicit provision in the 1990 Amendments to the Clean Air Act calls for a comprehensive benefit-cost assessment of the 1990 amendments. The EPA has made little progress toward building a database for that assessment and is now making decisions under Title III that will make such an assessment very difficult to perform in the future.

The missteps of the EPA under Title III demonstrate why it is important for Congress to pass an across-the-board statutory requirement that all risk-protection rules be accompanied by risk estimates and estimates of benefits and costs. Even if an agency's statutory mandate does not require or permit consideration of risk or benefit-cost comparisons, Congress and the public need to be informed of what risk regulation accomplishes and sacrifices. Otherwise, a future public debate about whether an agency's authority and resources should be contracted, retained, or expanded will have no factual basis.

It is desirable to prevent excessive analytical burdens, but Congress should not accomplish that by simply exempting regulations that cost less than $100 million per year. Such a move creates perverse incentives for rules to be subdivided into smaller components and for cost estimates to be trimmed to come in under the $100 million threshold. Thus, Congress should simply authorise agencies to tailor the complexity of the analysis to the importance of the rule. Reasonable Relationship between Cost and Risk Reduction

In the 1960s and 1970s Congress was understandably hesitant to require agencies to show that every risk regulation has marginal monetary benefits in excess of marginal monetary costs. Even today, despite twenty-five years of progress in the

3 Interview with Mr. Barry Felrice, Office of Rulemaking, National Highway and Traffic Safety Administration, Washington, D.C., 1994.

16 EMF Risk Perception and Communication science of quantifying and monetising risks, it would not be wise to impose a strict net-benefit test on rulemaking.

Many human health and environmental benefits remain difficult to quantify (for example, the monetary value of slightly improved visibility on summer days). Progress has been made in methods of "contingent valuation" of health and environmental benefits, but substantial obstacles remain in the confident application of those tools to risk regulation (Freeman 1993). At the same time, many of the more subtle economic effects of rules are difficult to quantify (such as the indirect impacts on industrial productivity and innovation). In addition, fairness and justice considerations may persuade us to adopt some rules that would "flunk" a strict net- benefit test. For instance, a proposed rule that does not satisfy a net-benefit test but promises a significant reduction in the risks and costs incurred by low-income and minority populations may be worth adopting on equity grounds.

While a strict net-benefit test is ill-advised, Congress should require agencies to make a plausible case that the benefits of a rule (quantitative and qualitative) bear a reasonable relationship to costs (quantitative and qualitative). Reasonable relationship is intended here as an intuitive narrative standard rather than a specific mathematical balance (Portney 1990). Of course, what one person perceives to be intuitively reasonable may seem nutty to someone else. Various federal agencies are clearly operating under different norms about what kinds of investments in risk reduction are reasonable.

Consider, for example, a little-noticed discrepancy between FDA and EPA decision making. Soon after the onset of the AIDS epidemic, the FDA ordered routine testing of the blood supply to protect recipients of donated blood from contracting the virus. Recently, the FDA considered and rejected the option of adding a special test for the immunodeficiency virus antigen (HIV-Ag) to prevailing procedures for testing donated blood for HIV. The estimated cost of applying that special test to all donated blood in the United States is $48 million per year. The estimated health benefit would be two to four fewer cases of AIDS per year, or a cost-effectiveness ratio of $12 million to $24 million per case of AIDS averted (Gelles 1993).4

If the EPA had proposed the same investment, it would very likely have been considered reasonable. Consider, for example, the EPAs final rule requiring

4 Note that those calculations exclude the epidemic control benefits resulting from fewer infected citizens.

17 EMF Risk Perception and Communication the reformulation of gasoline in an effort to reduce human exposures to carcinogens such as benzene, a known human leukemogen, and volatile organic compounds, which are precursors to smog. The estimated annual costs of that rule are $700 million per year in Phase I plus an additional $250 million in Phase II. The esti- mated benefits of Phase I are twenty fewer cases of cancer plus 115,000 fewer tons of volatile organic compounds emitted into the atmosphere. Phase II benefits are estimated to be four fewer cancer cases, 42,000 fewer tons of volatile organic compounds, and 22,000 fewer tons of nitrogen oxides. Those benefit estimates are based on assumptions that are far more likely to exaggerate benefits than underestimate them.

If we make the further assumption that each ton less of volatile organic compound or nitrogen oxide produces $5,000 in general health and ecological benefits (which the EPA regards as generous), then the net benefits of Phases I and II (excluding the cancers avoided) are - $125 million and - $140 million, respectively. The resulting incremental cost-effectiveness ratios are $6.25 million and $35 million per case of cancer avoided (EPA 1994d).

This comparison of FDA and EPA decisions suggests that either the government values the lives of patients receiving donated blood less than the lives of potential cancer patients or that the FDA and the EPA are not behaving consistently. I tend to favour the latter conclusion, since in a recent study covering hundreds of medical and environmental health policies, we found that federal regulators frequently make investments in toxin control that would not be considered reasonable by the norms of preventive medicine (Tengs, Adams, Pliskin, Gelb- Safran, Siegel, Weinstein, and Graham 1995). Congress should address that discrepancy by requiring agencies to achieve a reasonable and consistent balance between benefits and costs. The OMB should be required to develop formal guidance on what magnitudes of investments in health protection are likely to be supportable by reference to the burgeoning literature on the public's willingness to pay for health protection (Viscusi 1992). By applying the same guidance to all agencies, the OMB would take a step toward effecting consistency in norms. External Mechanisms of Scientific Peer Review

Since analysts wield a subtle yet important power when conducting various types of risk analysis, Congress should insist that their analytical products be scrutinised and improved through the external mechanism of peer review (Burack 1987; Graham 1991a). A useful model for review, one the EPA and FDA often use, is a public advisory committee of non-governmental scientists from academia and

18 EMF Risk Perception and Communication non-profit research organizations. Members are selected on the basis of their technical expertise rather than on the basis of their affiliation with particular stakeholder groups. Advisory committees may consider comments from stakeholders in a public forum, scrutinise a draft agency report, and provide written technical advice to an agency (Lippman 1987). Studies have shown that a public process of external peer review improves both the quality of the technical analysis and the degree of public confidence in decisions that are ultimately made based on the analysis (Jasanoff 1990).

When writing peer review requirements into risk legislation, Congress should insist that such review be applied to economic and engineering analyses as well as to public health and ecological analyses. The extent of peer review should match the importance and complexity of the issue.

Some agencies, such OSHA, NHTSA, and the CPSC, do not yet make widespread use of external advisory groups. The EPA has an improving track record in that area (EPA annual), but serious problems remain. The EPAs Integrated Risk Information System, which supplies If Congress believes in risk sensitive information on cancer and analysis, it should not be bashful noncancer effects in computerised form about authorizing judicial review of to the public, has not yet been subjected an agency’s use of risk analysis in to a rigorous yet flexible mechanism of rulemaking decisions. peer review. Problems also occur when agencies such as the EPA or OSHA accept a "science-policy" determination made by an international organization without independent peer review. For example, the World Health Organization's International Agency for Research on Cancer categorises chemicals and processes as “carcinogenic” without considering information on biological mechanisms of action and extent of human exposure. Congress should insist that federal agencies consider the recommendations of international organizations such as IARC, but only in conjunction with independent review by a public advisory committee of scientists in the United States.

Judicial Review under the Principle of Deference

Were agencies already making optimal use of risk analysis, legislation would not be urgent. The country needs legislation precisely because the rate of progress in the analytic practices of agencies is uneven and often slow (Stone 1994; Landy, Roberts, and Thomas 1990). Legislation can spur agency activities, but only

19 EMF Risk Perception and Communication if parties outside the agency have the opportunity to bring the agency to court for failure to do so.

If Congress believes in risk analysis, it should not be bashful about authorising judicial review of an agency's use of risk analysis in rulemaking decisions under a deferential standard of review such as the "arbitrary and capricious" test in the Administrative Procedures Act and subsequent case law (Breyer 1982). Congress should not authorise judges to substitute their scientific or policy judgements for those of the agency (Graham, Green, and Roberts 1988). Risk analysis is unlikely to influence administrative decision making if decision makers are not compelled to consider seriously the findings of analysis when making decisions (Melnick 1983). Legislative reform can be influential in addressing the misallocation of resources resulting from the public's current syndrome of paranoia and neglect about risk (Cross 1994).

Analytical Resources

The congressional commitment to risk analysis should include more budgetary and technical resources as well as more analytical requirements. A recent report by the OTA (1994) documented the minimal resource commitments in the federal government that have been made to advancing the field. Supreme Court justice Stephen Breyer (1993) has also highlighted the need to cultivate and support a cadre of career public servants who have broad multidisciplinary experience in risk analysis. The need for analytical resources is pressing in all fields but especially in the sub-field of ecological risk assessment - the most immature aspect of this growing discipline (Bamthouse 1994).

Looking much more broadly, it is increasingly apparent that the concept of risk needs to be integrated into the way scientists and professionals are trained. Otherwise, the needs of the public and private sectors for experts and keen decision makers will not be met. In the long run, the nation's commitment to risk analysis needs to be expressed not only in the education of scientists and professionals but in the curricula used to educate young children in math, science, and economics. Perhaps this more analytical citizenry is the ultimate cure for the prevailing syndrome of paranoia and neglect about risk.

NOTES

20 EMF Risk Perception and Communication

Helpful suggestions were made by John Evans, Robert Hahn, James Hammitt, Jennifer Hartwell, March Sadowitz, Paul Slovic, and Jonathan Wiener. The views are exclusively those of the author.

REFERENCES

Abelson, Philip H. "Pesticides and Food." Science 259 (1993): 1235.

Ames, Bruce N. "Pollution, Pesticides, and Cancer." Journal of AOAC International 75 (1992): 1-5.

------, M. Profet, and Lois S. Gold. "Dietary Pesticides (99.9% All Natural)." Proceedings of the National Academy of Sciences, USA 87 (1990): 7777-81.

Archibald, S. 0., and C. K. Winter. "Pesticides in Our Food: Assessing the Risks," in C. K. Winter, J. N. Seiber, and C. F. Nuckton, eds., Chemicals in the Human Food Chain. New York: Van Nostrand Reinhold, 1990.

Barnthouse, L. W. "Issues in Ecological Risk Assessment: The CRAM Perspective." Risk Analysis 14 (1994): 251-56.

Brain, J. D. 9 B. D. Beck, A. J. Warren, and R. A. Shaikh, eds. Variations in Susceptibility to Inhaled Pollutants: Identification, Mechanisms, and Policy Implications. Baltimore: Johns Hopkins University Press, 1988.

Breyer, Stephen G. Breaking the Vicious Circle. Cambridge: Harvard University Press, 1993.

------. Regulation and Its Reform. Cambridge: Harvard University Press, 1982.

Burack, T. S. "Of Reliable Science: Scientific Peer Review, Federal Regulatory Agencies, and the Courts." Virginia Journal of Natural Resources Law 7 (1987): 27-110.

Business Roundtable. Toward Smarter Regulation. Washington, D.C.: Business Roundtable, 1994.

21 EMF Risk Perception and Communication

Carmegie Commission on Science, Technology, and Government. Risk and the Environment: Improving Regulatory Decision Making. New York: Carnegie Commission, June 1993.

Considine, T.J., G. A. Davis, and D. Marakovits. "Costs and Benefits of Coke Oven Emission Controls." Final Report to U.S. EPA, December 17, 1992.

Cooke, R. M. Experts in Uncertainty: Opinion and Subjective Probability in Science. New York: Oxford University Press, 1991.

Cross, F. "The Public Role in Risk Control." Environmental Law 24 (1994): 888- 969.

Department of Health and Human Services. Health USA. Washington, D.C.: Government Printing Office, annual.

------. Healthy People: National Health Promotion and Disease Prevention Objectives. Washington, D.C.: Government Printing Office, September 1990.

Drummond, M. F., G. L. Stoddart, and G. W. Torrance. Methods for the Economic Evaluation of Health Care Programmes. New York: Oxford University Press (Oxford Medical Publications), 1987.

Dunlap, R. E. "Trends in Public Opinion toward Environmental Issues: 1965-1990." Society and Natural Resources 4 (1991): 285-312.

Eddy, D. M. "Screening for Cervical Cancer." Annals of Internal Medicine 113 (1990): 214-226.

Environmental Law Review Committee. Environmental Legislation. The Increasing Cost of Regulatory Compliance to the City of Columbus, Report to the Mayor and City Council of Columbus, Ohio. May 13, 1991.

Environmental Protection Agency. The Dioxin Reassessment. External Review Draft. Washington, D.C., June 1994a. ------. A Guidebook to Comparing Risks and Setting Environmental Priorities. Washington, D.C.: Government Printing Office, September 1993b.

22 EMF Risk Perception and Communication

------. "National Emission Standard for Hazardous Air Pollutants for Source Cate- gories: Organic Hazardous Air Pollutants from the Synthetic Organic Chemical Manufacturing Industry and Other Processes Subject to the Negotiated Regula- tion for Equipment Leaks." 40 CFR Part 63, 1994c.

------. "National Emission Standards for Hazardous Air Pollutants; Benzene; Rule and Proposed Rule." Federal Register 54 (September 14, 1989): 38044-72.

------. "National Emission Standards for Hazardous Air Pollutants for Source Categories and for Coke Oven Batteries." Federal Register 58 (1993a): 57898- 935.

------. "National Emission Standards for Hazardous Air Pollutants for Source Categories: Aerospace Manufacturing and Rework." Federal Register 59 (1994b): 29216-51.

------. "Regulation of Fuels and Fuel Additives: Standards for Reformulated and Conventional Gasoline." Federal Register 59 (February 16, 1994d).

------. Report of the Director of the Science Advisory Board. Washington, D.C.: Government Printing Office, annual.

Environmental Protection Agency, Science Advisory Board. Reducing Risk. Washington, D.C.: Government Printing Office, 1990.

Federal Focus Inc. Toward Common Measures: Recommendations for a Presidential Executive Order on Environmental Risk Assessment and Risk Management Policy. Washington, D.C.: Federal Focus Inc., 1991.

Fischhoff, Baruch. "Acceptable Risk: A Conceptual Proposal." Risk: Health, Safety and Environment 5 (1994): 1-28.

------, et al. "How Safe is Safe Enough? A Psychometric Study of Attitudes toward Technological Risks and Benefits." Policy Sciences 8 (1978): 127.

Fischhoff, B., S. Lichtenstein, P. Slovic, S. L. Derby, and Ralph L. Keeney. Acceptable Risk. New York: Cambridge University Press, 1981.

23 EMF Risk Perception and Communication

Florini, K. L., and E. K. Silbergeld. "Getting the Lead Out." Issues in Science and Technology 9 (Summer 1993): 33-39.

Foster, K. R., D. E. Bernstein, and Peter W. Huber, eds. Phantom Risk: Scientific Inference and the Law. Cambridge: MIT Press, 1994.

Freeman, A. M. The Measurement of Environmental and Resource Values: Theory and Methods. Washington, D.C.: Resources for the Future, 1993.

Gelles, G. M. "Costs and Benefits of HIV-1 Antibody Testing of Donated Blood." Journal of Policy Analysis and Management 12 (1993): 512-31.

Goldstein, Bernard D., M. Demak, M. Northridge, and D. Wartenberg. "Risk to Groundlings of Death Due to Airplane Accidents: A Risk Communication Tool." Risk Analysis 12 (1992): 339-41.

Graham, John D. "The Economics of Controlling Outdoor and Indoor Air Pollution," in L. Tomatis, ed., Indoor and Outdoor Air Pollution and Human Cancer. New York: Springer-Verlag, 1993.

------. New Directions for Chemical Risk Assessment." Regulation (Fall 1991b): 14-18.

------. The Safety Risks of Fuel Economy Legislation." Risk: Issues in Health and Safety 3 (1992): 95-126.

------, ed. Harnessing Science for Environmental Regulation. Westport, Conn.: Greenwood Publishing Co., 1991a.

Graham, John D., and J. K. Hammitt. "Refining Comparative Risk Assessment for Priority Setting in the Federal Government," in Terry Davies, ed., Comparative Risk. Washington, D.C.: Resources for the Future, in press.

Graham, John D., and Jonathan Wiener. Risk versus Risk: Tradeoffs in Human Health and Environmental Protection. Cambridge: Harvard University Press, 1995. Graham, John D., L. Green, and Marc J. Roberts. In Search of Safety: Chemicals and Cancer Risk. Cambridge: Harvard University Press, 1988.

24 EMF Risk Perception and Communication

Hegner, R. E. "Does Protecting for Human Health Protect Ecological Health?" Risk Analysis 14 (1994): 3-4.

Jasanoff, Sheila. The Fifth Branch: Science Advisors as Policymakers. Cambridge: Harvard University Press, 1990.

Laden, F., and G. M. Gray. "Toxics Use Reduction: Pro and Con." Risk: Issues in Health and Safety 4 (1993): 213-34.

Landy, M. K., Mark J. Roberts, and S. R. Thomas. EPA: Asking the Wrong Questions. New York: Oxford University Press, 1990.

Larson, J. L., D. C. Wolf, and B. E. Butterworth. "Induced Cytotoxicity and Cell Proliferation in the Hepatocarcinogenicity of Chloroform in Female B6C3F1 Mice: Comparison of Administration by Gavage in Corn Oil vs. ad Libitum in Drinking Water." Fundamental and Applied Toxicology 22 (1994): 90-102.

Lippman, M. "Role of Science Advisory Groups in Establishing Standards for Ambient Air Pollutants." Aerosol Science and Technology 6 (1987): 93-114.

McGinnis, J. M., and W. H. Foege. "Actual Causes of Death in the United States." Journal of the American Medical Association 270 (1993): 2207-12.

Melnick, R. Shep. Regulation and the Courts: The Case of the Clean Air Act. Washington, D.C.: Brookings Institution, 1983.

Mendeloff, J. M., and R. M. Kaplan. "Are Large Differences in 'Lifesaving' Costs justified? A Psychometric Study of the Relative Value Placed on Preventing Deaths." Risk Analysis 9 (1989): 349-63.

Merkhofer, M. W. Decision Science and Social Risk Management. Boston: D. Reidel Publishing Company, 1987.

Morgan, M. G., and M. Henrion. Uncertainty: A Guide to Dealing with Uncertainty in Quantitative Risk and Policy Analysis. New York: Cambridge University Press, 1991. Municipality of Anchorage. Paying for Federal Environmental Mandates: A Looming Crisis for Cities and Counties. Anchorage: September 1992.

25 EMF Risk Perception and Communication

National Committee for Injury Prevention and Control. Injury Prevention: Meeting the Challenge. New York: Oxford University Press, 1989.

National Highway Traffic Safety Administration. Development of the Functional Capacity Index. Washington, D.C.: Department of Transportation, July 1994.

National Research Council. Environmental Epidemiology: Public Health and Hazardous Wastes. Washington, D.C.: National Academy Press, 1991.

------. Human Exposure Assessment to Airborne Pollutants. Washington, D.C.: National Academy Press, 1991.

------. Issues in Risk Assessment. Washington, D.C.: National Academy Press, 1992.

------. Science and judgement in Risk Assessment. Washington, D.C.: National Academy Press, January 1994.

------. The Structure and Performance of the Health Effects Institute. Washington, D.C.: National Academy Press, 1993.

Office of Technology Assessment. Health Risk Assessment. Washington, D.C.: Government Printing Office, 1994.

Portney, Paul R. "Economics and the Clean Air Act." Journal of Economic Perspectives 4, (Fall 1990): 173-81.

------, ed. Public Policies for Environmental Protection. Washington, D.C.: Resources for the Future, 1990.

Roth, E., M. G. Morgan, B. Fischhoff, Lester Lave, and A. Bostrom. "What Do We Know about Making Risk Comparisons?" Risk Analysis 10 (1990): 375-87.

Samet, J. M., and J. D. Spengler, eds. Indoor Air Pollution: A Health Perspective. Baltimore: Johns Hopkins University Press, 1991.

Singer, Eleanor, and Phyllis M. Endreny. Reporting on Risk. New York: Russell Sage Foundation, 1993.

26 EMF Risk Perception and Communication

Stone, Richard. "Can Carol Browner Reform EPA?: One Year after Becoming EPA Administrator, Browner Has a Lot to Do to Fulfill Her Promise to Make Science the Centerpiece of Environmental Regulation." Science 263 (1994): 312-14.

Tengs, Tammy 0., Miriam E. Adams, Joseph S. Pliskin, Dana Gelb-safran, Joanna E. Siegel, Michael C. Weinstein, and John D. Graham. "Five-Hundred Life- Saving Interventions and Their Cost-Effectiveness." Risk Analysis 15 (1995): 369-90.

Viscusi, W. Kip. Fatal Tradeoffs: Public and Private Responsibilities for Risk. New York: Oxford University Press, 1992.

Warren, M. "Reforming the Federal Regulatory Process: Rhetoric or Reality?" Working Paper, Center for the Study of American Business, Washington University, June 1994.

Wilson, R. "Analyzing the Risks of Everyday Life." Technology Review 81 (1979): 40-46.

Zeckhauser, Richard J., and W. Kip Viscusi. "Risk within Reason." Science 238 (1990): 559-64.

Zimmerman, Rae. "Social Equity and Environmental Risk." Risk Analysis 13 (1993): 649-66.

27

EMF Risk Perception and Communication

Presentations

EMF concerns and WHO’s International EMF Project

Dr. M.H. Repacholi World Health Organization, CH-1211 Geneva 27, Switzerland Tel: +44 21 791-3427, Fax: +44 21 791-4123 E-mail: [email protected]

1. Introduction

The World Health Organization takes seriously the concerns raised by reports about possible health effects from exposure to electromagnetic fields (EMF). Cancer, changes in behaviour, memory loss, Parkinson and Alzheimer's diseases, and many others have been suggested as resulting from exposure to EMF. Everyone in the world is now exposed to a complex mix of EMF frequencies. EMF has become one of the most pervasive environmental influences and exposure levels at many frequencies are increasing significantly as the technological revolution continues unabated and new applications using different parts of the spectrum are found. Major sources of EMF exposure include electric power generation, distribution and use, transportation systems, telecommunications facilities and associated devices such as mobile telephones, medical and industrial equipment, radar and radio and television broadcast antennas.

WHO established the International Electromagnetic Fields (EMF) Project to assess health and environmental effects of exposure to static and time varying electric and magnetic fields in the frequency range 0 - 300 GHz. The Project commenced at WHO in 1996. It has been designed to follow a logical progression of activities and produce a series of outputs to allow improved health risk assessments to be made and to identify any environmental impacts of EMF exposure. The ultimate objective of the Project is to provide the health risk assessments that will lead to the development of an international consensus on exposure guidelines.

How people perceive risks, understanding concerns expressed by the general public and workers, and how scientists, industry and governments can more effectively deal with these concerns, is essential to our overall understanding of all facets of the EMF issue. This paper reviews possible consequences to health from

29 EMF Risk Perception and Communication

EMF exposure and summarises what WHO is doing to address this issue. Details on the EMF Project are available on the home page at: http://www.who.ch/emf/.

International EMF Project

WHO established the International EMF Project to provide a mechanism for resolving the many and complex issues related to possible health effects of EMF exposure. The Project assesses health and environmental effects of exposure to static and time varying electric and magnetic fields in the frequency range 0 - 300 GHz, with a view to the development of international guidelines on exposure limits. It commenced at WHO in 1996 and will end in 2005.

The EMF Project has been designed in a logical progression of activities and outputs to allow improved health risk assessments to be made, and to identify any environmental impacts of EMF exposure. The Project objectives are to:

1. Provide a co-ordinated international response to the concerns about possible health effects of exposure to EMF,

2. Assess the scientific literature and make a status report on health effects,

3. Identify gaps in knowledge needing further research to make better health risk assessments,

4. Encourage a focused research programme in conjunction with research funding agencies,

5. Incorporate the research results into WHO's Environmental Health Criteria monographs where formal health risk assessments will be made on exposure to EMF,

6. Facilitate the development of an international consensus on limits for EMF exposure,

7. Provide information on the management of EMF protection programmes for national and other authorities, including monographs on EMF risk perception, communication and management, and

30 EMF Risk Perception and Communication

8. Provide advice to national authorities, other institutions, the general public and workers, about any hazards resulting from EMF exposure and any needed mitigation measures.

WHO, through its International EMF Project, has recently conducted in- depth international reviews of the scientific literature on the biological and health effects of exposure to radiofrequency (RF), and static and extremely low frequency (ELF) fields. These reviews were conducted with the purpose of identifying;

1. health effects that can be substantiated from the literature, and

2. biological effects that are suggestive of possible health effects, but require further research to determine if exposure to electromagnetic fields (EMF) at the low levels of exposure normally encountered in the living and working environment has any impact on health.

The results of these reviews have been or are being published in the Bioelectromagnetics journal. Research still needed to fill these gaps in knowledge form the WHO EMF Research Agenda that is available on the EMF Project home page (http://www.who.ch/emf/) or from WHO.

Having completed the initial international scientific reviews, WHO is now urging EMF funding agencies world wide to give priority to this research, if it is their intention to obtain results that will assist both WHO and the International Agency for Research on Cancer (IARC) to make better health risk assessments.

Both WHO and IARC have already established a timetable for assessing health effects of EMF fields. In 2001 IARC will conduct a meeting to formally identify and evaluate the evidence for carcinogenesis from exposure to static and extremely low frequency (ELF) fields. IARC will publish the results of this meeting in the IARC Monograph Series. The International EMF Project will accept the IARC conclusions on carcinogenesis and incorporate them into the results of a WHO evaluation of non-cancer health risk assessment of exposure to static and ELF fields in 2002. The results and conclusions will be published in the Environmental Health Criteria series. It is anticipated that sufficient results will be available for IARC to conduct a similar evaluation of evidence for carcinogenicity of RF fields in 2003. WHO would then complete an overall health risk assessment of exposure to RF fields in 2004.

31 EMF Risk Perception and Communication

Biological Effects of EMF Fields

The following provides a summary of the results of international reviews of the biological and health effects of exposure to low levels of RF fields, held in Munich in November 1996, and of exposure to static and ELF fields, held in Bologna in June 1997. References to the original work can be found in the WHO reviews noted in the “Further Reading” section. Conclusions from the review conducted by the National Institute of Environmental Health Sciences (NIEHS) under its RAPID program are also summarised.

2. ELF Fields

Interaction mechanisms

A well-known mechanism of interaction of ELF fields with biological tissues is the induction of time-varying electric currents and fields. At sufficiently high levels, these can produce direct stimulation of excitable tissues such as nerve and muscle cells. At the cellular level, the interaction induces voltages across the membranes of cells sufficient to stimulate nerves to conduct or muscles to contract. This mechanism accounts for the ability of humans and animals to perceive electric currents in their bodies and to experience electric shocks. Other mechanisms have been proposed, but there is little evidence to support them.

Electric fields

External ELF electric fields induce time-varying electric charges on the surface of the body. The magnitude and distribution of the charges depend on the body shape and its location and orientation relative to the field and ground plane. In addition, electric fields, electrical polarisation changes, and currents are induced inside the body as a result of time-variation of this surface charge density. Charges fixed on internal molecules polarise and depolarise as the field changes. Since time- variation in the ELF range is slow compared to the ability of charges to move, the fields and currents generated inside the body from this source are very small. The induced current density distribution depends on the electrical properties of the tissue and varies inversely with the body cross-section. Typically, the strength of the internal electric fields is less than 10-6 of the external field.

Magnetic fields

32 EMF Risk Perception and Communication

The induced current density is proportional to the rate of change of the magnetic flux density. For sinusoidal applied fields, the induced fields and currents are linearly dependent on frequency. The magnitude of the currents induced by pulsed magnetic fields will depend on the rise and fall time of the pulse. The highest current densities are induced in peripheral tissues, since these have the largest inductive loop radius in the body. However, tissue inhomogeneity and orientation of the body to the field will affect the current path. In general, the electric field induced in peripheral tissues by a horizontal magnetic field is approximately 1.5 times that induced by a vertical magnetic field of similar magnitude. Currents circulating from head to foot due to a horizontal magnetic field will be high in the neck because its small cross-section concentrates the flow.

For a human with torso radius of 0.15 m and tissue conductivity of 0.2 S/m, a 50 Hz magnetic field parallel to the long axis of the body will induce a current in the tissue periphery of about 5 A/m2 per tesla. Since current density is proportional to body radius, current density values can be used to scale between animal and human exposure. Typical induced currents and fields for 1 µT, 60 Hz uniform magnetic field-exposure of mice, rats and humans are in the range of 0.1-0.4, 0.3- 1.3, and 1-20 µA/m2, respectively.

ELF biological effects

Laboratory studies

Above about 0.1 mT, a variety of studies have demonstrated effects in vitro on ornithine decarboxylase (ODC) activity. Not all replication attempts have succeeded, however. Many other biological effects have been reported above about 1 mT. How magnetic field exposure produces such effects is unknown. For most effects, such as those reported on genotoxicity, intracellular calcium concentrations, or general patterns of gene expression, convincing and reproducible results have not been observed. None of the in vitro effects are necessarily indicative of an adverse health effect. Without knowledge of the mechanisms involved, effects observed at high field strengths cannot be extrapolated to lower fields, since the mechanisms may be different.

While there is no convincing evidence that ELF fields cause cancer in animals, only a limited number of studies have been conducted to test this hypothesis. Some recent studies suggest a positive relationship between breast cancer in animals treated with carcinogens and ELF magnetic field exposure at

33 EMF Risk Perception and Communication approximately 0.02-0.1 mT. The importance of these findings needs to be investigated further. Currently available data do not provide convincing evidence of adverse effects from exposure to power frequency fields on reproduction or development in mammals. There is evidence of behavioural and neurobehavioural responses in animals, but only following exposure to strong 60 Hz electric fields.

Neuroendocrine changes are associated with exposure to ELF magnetic fields, but these alterations have not been shown to cause adverse effects in animals. Some studies suggest magnetic fields of strength between 0.01 and 5.2 mT might inhibit night-time pineal and blood melatonin concentrations in experimental animals. However, such effects have not been demonstrated in humans.

Human laboratory studies

Perception

Exposure to ELF electric fields can result in field perception as a result of alternating electric charge induced on the surface causing body hair to vibrate. Most people can perceive electric fields greater than 20 kV/m, and a small percentage of people perceive field strengths below 5 kV/m. In two well-controlled studies, humans were unable to perceive magnetic fields at levels up to 1.5 mT.

During exposure to ELF magnetic fields above 3-5 mT, volunteers experience faint visual flickering sensations or magnetophosphenes. The threshold current density in the retina for induction of magnetophosphenes is about 10 mA/m2 at 20 Hz, well above typical endogenous current densities in electrically excitable tissues. Higher thresholds have been observed for both lower and higher frequencies.

Cardiovascular

Several reports indicate that ELF fields influence the cardiovascular system. Exposure of human volunteers to combined 60 Hz electric and magnetic fields (9 kV/m, 0.02 mT) resulted in small changes in cardiac function. Resting heart rates were found to be slightly but significantly reduced (about 3-5 beats/minute) during or immediately after exposure. This response did not occur with exposure to stronger (12 kV/m, 0.03 mT) or weaker (6 kV/m, 0.01 mT) fields and was reduced if the subject was mentally alert. In these double-blind studies, subjects were unable to detect the presence of the fields. While continuous exposure to combined electric

34 EMF Risk Perception and Communication and magnetic fields at 9 kV/m, 0.02 mT slows the heart, intermittent exposure can result in both slowing and increasing heart rate. None of the effects on heartbeat exceeded the normal range. No obvious acute or long-term cardiovascular-related hazards have been demonstrated at levels below current exposure standards for ELF or radio frequency fields.

Spectral analysis of electrocardiograms indicated that intermittent exposure to ELF reduced power in the Fourier spectrum at frequencies associated with blood pressure and thermoregulatory control mechanisms, and increased the power associated with respiration. These results were replicated in another sample of volunteers using a different experimental design. In contrast, a third study by the same group failed to show any effects on heart rate variability when subjects were exposed continuously rather than intermittently to the same ELF field. The patterns of heart rate variability observed in the above studies are similar, but not identical to, the pattern found to be predictive of sudden cardiac death.

Hormone and Immune System Effects

No changes in blood chemistry, blood cell count, blood gases, lactate concentration, skin temperature or circulating hormones have been observed. Field- related suppression of the hormone melatonin has been proposed as a mechanism for the relationship between exposure to magnetic fields and increased cancer risk reported alterations in melatonin production in people sleeping at home under electric blankets, although all subjects did not show melatonin suppression. Suppression of melatonin has also been reported in studies of Finnish garment workers, electric utility workers, and VDU workers. In contrast, well-controlled laboratory studies report mostly negative results. Some laboratory studies, however, have reported positive results.

Two recent French reports indicate that acute exposure for one night to a linearly polarised magnetic field at 10 µT has no effect on hormonal or immune parameters in healthy male volunteers. No published reports have examined possible differential effects in women, possible influence of longer exposure or of altering field polarisation.

Epidemiological studies

Residence in homes near external power lines is associated with an approximate 1.5-fold relative risk of childhood leukaemia. Although the literature is

35 EMF Risk Perception and Communication inconsistent, when studies using various markers of exposure, such as proximity to power lines and calculated magnetic fields from power lines, are combined, the increased risk is statistically significant (NRC, 1996). While confounding factors and reporting or case selection bias may have influenced some of the studies, they are unlikely to account for the overall pattern of association. The recent US National Cancer Institute study weakens these associations in meta-analyses, but the association between measured field strength and leukaemia still remains statistically significant. The association between wire codes and childhood leukaemia has not been explained, although in most studies average magnetic field levels measured in the homes of children are less closely associated than wire codes with an excess in childhood leukaemia or any other cancers. External wiring is not a strong predictor of magnetic field exposure, and so other factors which are associated with the wire code, but which may or may not be field-related, may explain the association (NRC, 1996).

Studies examining health outcomes other than childhood cancer do not provide sufficient evidence to support an association between ELF magnetic-field exposure and adult cancers, pregnancy outcome, or neurobehavioural disorders (NRC, 1996).

Occupational studies have generally used job titles, sometimes in combination with workplace ELF field measurements, to determine if any association exists between exposure to these fields and cancer. Elevated risks of various cancers have been reported, especially leukaemia, nervous system tumours and breast cancer; but the lack of uniformity of the results has been a major concern. Any excess cancer risk among electrical workers, compared to other occupations, is small and difficult to detect using epidemiological methods. Studies so far have been complicated by the lack of adequate exposure assessment in the workplace and possible confounding factors.

The basic problem with all epidemiological studies so far has been the lack of any concept of dose or an exposure metric established from laboratory studies. Metrics used have generally been cumulative exposure or time-weighted average field strength. Very little information has been obtained about exposure from appliances, ground currents or devices that may be associated with transient fields. Brief exposures to high-amplitude magnetic field transients or to high-frequency harmonics have not been assessed in published studies. Personal dosimeters do not exist that can capture this information.

36 EMF Risk Perception and Communication

NIEHS Working Group conclusions on cancer

The following are the concluding remarks from the report of a Working Group formed by the NIEHS (1998) to evaluate the health effects from exposure to ELF. The Working Group concluded that ELF fields are a possible human carcinogen. The evidence in support this decision resulted from studies on childhood leukaemia in residential environments and on chronic lymphocytic leukaemia (CLL) in adults in occupational settings. “The fact that limited evidence was seen for CLL in adults should not be construed as providing support for the finding with regard to leukaemia in children. Childhood leukaemia and adult CLL are very different diseases with different aetiologies. Also, the inadequacy of the evidence for an effect on the risk for CLL in adults in the studies of residential exposure neither supports nor refutes the findings in the studies of occupational exposure. The in-vitro and mechanistic data provide, at best, marginal support for the conclusion that ELF fields are possibly carcinogenic to humans. While ELF magnetic fields at intensities greater than 100 µT provide moderate support for effects in vitro, there was little evidence of effects at intensities below this limit, which cover most of the range of exposure in the studies of residential childhood exposure and adult occupational exposure. Relatively few of the studies of occupational exposure addressed exposure to electric fields. Finally, the inadequate evidence from long-term bioassays for carcinogenicity in rodents is driven more by lingering concerns about single findings in two separate studies than by an overall concern that something has been missed in these studies or that there is a trend toward a positive effect in poorly conducted studies.”

The NIEHS Working Group’s conclusion suggests that further focused research is needed to clarify the possibility that ELF is carcinogenic. Research suggested by WHO to fill current gaps in knowledge is given below.

ELF Research needs

Independent replication of some key studies is a high priority. When effects are robust, replication should be straightforward and can be used as a basis for extending observations. It is important to characterise the dose-response relationship (field strength, threshold and exposure duration) of any effect, particularly at environmentally relevant field strengths.

Where possible, in vivo studies should consider exposures that include intermittency, transients, and duration as important variables. In addition, it would

37 EMF Risk Perception and Communication be valuable to consider the interactions of ELF fields with other agents, such as ionising radiation and chemicals. These interactions should test the hypothesis that ELF fields may act as a co-promoter for cancer, but other end points suggested by the in vitro literature should also be examined. Wherever possible, exposures should be relevant to those experienced by humans in occupational and residential settings. Some cancer-related studies using various animal models are currently under way. Research gaps for which additional results are needed, are as follows:

1. Confirmation and extension of animal studies reporting increased tumour incidence when magnetic fields are applied in combination with chemical carcinogens. These experiments should focus on dose- response relationships and the relationship between different exposure conditions.

2. Confirmation and extension of studies suggesting that magnetic field exposure influences mammary cancer development. Possible changes in relevant hormonal factors in magnetic field-exposed animals and controls should be investigated to examine potential mechanisms.

3. Neurophysiology/neurobehavioural studies using models of neurodegenerative diseases are indicated because of recent reports of possible ELF-field influence on human neurodegenerative diseases, such as Alzheimer disease.

4. While most studies of ELF field effects on various end points in reproduction and development have been negative, new studies should provide information on long-term neurobehavioural consequences following in utero exposure to magnetic fields. These studies should address whether ELF fields can produce effects on early brain development as measured in functional behaviour in adult animals.

Epidemiology

The most important prerequisite for future epidemiological studies is a clearer understanding of what metric should be used to characterise ELF field exposure. This may come from laboratory work or from additional hypothesis- generating epidemiological studies, each of which has advantages and disadvantages in cost, time, and precision. Project designs for new epidemiological research should, within the limits of what is possible, increase the role of measured past and

38 EMF Risk Perception and Communication present exposures. Dependence on surrogates, such as wire codes and job classifications, should decrease, particularly if data do not exist that establish how well the surrogates select for historical exposure. The a priori estimates of the power of future studies must be strong enough to predict useful information, given the outcomes of past research.

Because many, but not all studies show a small but significant excess in childhood leukaemia associated with residence in high wire code homes in the US (the only country where this surrogate has been used) a concerted effort is needed to explain this association. While efforts have been made to define the relationship between wire codes and average magnetic field exposure or socio-economic confounding factors, little evidence is available about the relationship between wire codes and high-amplitude transient fields or high-frequency harmonics. Future studies should include these and ground currents in the exposure assessments. Another aspect to be seriously pursued in future studies is the inclusion of non- occupational exposure.

With the above caveats, needed future epidemiological studies include:

1. Studies of the relationship between exposure and cancer incidence that properly assess both residential and occupational exposure over long periods, including transient magnetic-field exposure and high-frequency harmonics.

2. Studies to determine if correlates of wire codes, such as traffic density, age of home and sociodemographic characteristics of home occupants, can explain the statistical relationship between wire codes and childhood leukaemia.

3. Studies of the relationship between breast cancer and field exposure, including evaluation of both average field levels and of transients and high-frequency components and taking into account both occupational and non-occupational exposures.

4. Studies of the relationship between neurodegenerative disorders and field exposure, including evaluation of the role of average fields levels, transients and high-frequency components. Both occupational and non- occupational exposures should be considered.

39 EMF Risk Perception and Communication

5. Studies of the relationship between heart disease end points and exposure to ELF fields, including evaluation of the role of transient and high frequency components

6. and taking into account both occupational and non-occupational exposures.

Volunteer studies

Further studies are needed, especially using transient and high frequency components typical of environmental ELF fields, to determine:

1. Whether any component of the human melatonin hormone system is susceptible to ELF field exposure and, if so, the likely health consequence of this susceptibility.

2. Whether sleep disruption, changes in neurotransmitter metabolism, and learning and memory are associated with ELF field exposure.

3. The relationship between field exposure and slowing and variability in heart rate.

4. Whether electrophysiological indices of central nervous system activity and function are affected by ELF fields.

Subjective effects

Given the limited evidence, but widespread concern about subjective effects, more research is needed to determine:

1. Whether these health effects can be substantiated and can be related to EMF exposure.

2. Why people experiencing apparent hypersensitivity and attributing it to EMF exposure, cannot determine reliably whether the fields are on or off in laboratory tests. The current laboratory results should be extended, and their relevance clarified.

40 EMF Risk Perception and Communication

3. Radiofrequency Fields

Mechanisms of interaction

RF fields induce torques on molecules, which can result in displacement of ions from unperturbed positions, vibrations in bound charges (both electrons and ions), and rotation and reorientation of dipolar molecules such as water. These mechanisms, which can be described by classical electrodynamic theory, are not capable of producing observable effects from exposure to low-level RF fields, because they are overwhelmed by random thermal agitation. Moreover, the response time of the system must be fast enough to allow it to respond within the time period of the interaction. Both considerations imply that there should be a threshold (below which no observable response occurs) and a cut-off frequency (above which no response is observed). These thresholds would be expected to be present even in more refined models if they correctly take into account thermal noise and the kinetics of the system.

Exposure to electromagnetic fields at frequencies above about 100 kHz can lead to significant absorption of energy and temperature increases. In general, exposure to a uniform (plane-wave) electromagnetic field results in a highly non- uniform deposition and distribution of energy within the body, which must be assessed by dosimetric measurement and calculation. For absorption of energy by the human body, electromagnetic fields can be divided into four ranges:

1. frequencies from about 100 kHz to less than about 20 MHz, where absorption in the trunk decreases rapidly with decreasing frequency, and significant absorption may occur in the neck and legs;

2. frequencies in the range from about 20 MHz to 300 MHz, at which relatively high absorption can occur in the whole body, and to even higher values if partial body (e.g., head) resonances are considered;

3. frequencies in the range from about 300 MHz to several GHz, at which significant local, non-uniform absorption occurs;

4. frequencies above about 10 GHz, at which energy absorption occurs primarily at the body surface.

41 EMF Risk Perception and Communication

In tissue, SAR is proportional to the square of the internal electric field strength. Average SAR and SAR distribution can be computed or estimated from laboratory measurements. Values of SAR depend on the following factors:

1. the incident field parameters, i.e. the frequency, intensity, polarisation, and source–object configuration (near field or far field);

2. the characteristics of the exposed body, i.e. its size, internal and external geometry, and the dielectric properties of the various tissues;

3. reflection, absorption and scattering effects associated with the ground or other objects in the field near the exposed body.

When the long axis of the human body is parallel to the electric field vector, and under plane-wave exposure conditions (i.e. far-field exposure), whole-body SAR reaches maximal values. The amount of energy absorbed depends on a number of factors, including the size of the exposed body. "Standard Reference Man", if not grounded, has a resonant absorption frequency close to 70 MHz. For taller individuals the resonant absorption frequency is somewhat lower, and for shorter adults, children, babies, and seated individuals it may exceed 100 MHz. The values of electric field reference levels are based on the frequency-dependence of human absorption. In grounded individuals, resonant frequencies are lower by a factor of about 2 (ICNIRP, 1998).

RF biological effects

Reports from in vitro research indicate that low-level RF fields may alter membrane structural and functional properties that trigger cellular responses. It has been hypothesised that the cell membrane may be susceptible to low-level RF fields, especially when these fields are amplitude-modulated at ELF frequencies. At high frequencies, however, low-level RF fields do not induce appreciable membrane potentials. They can penetrate the cell membrane and possibly influence cytoplasmic structure and function. These RF field-induced alterations, if they occur, could be anticipated to cause a wide variety of physiological changes in living cells that are only poorly understood at the present time.

A lack of effects of RF exposure on mutation frequency has been reported in a number of test samples including yeast and mouse lymphoid cells. No effect of

42 EMF Risk Perception and Communication

RF-field exposure on chromosome aberration frequency in human cells has been confirmed.

In contrast to the evidence given above, several rodent studies indicate that RF fields may affect DNA directly. These papers report quantitative data subject to sources of inter-trial variation and experimental error such as incomplete DNA digestion or unusually high levels of background DNA fragmentation. These experiments need to be replicated before the results can be used in any health-risk assessment, especially given the weight of evidence that RF fields are not genotoxic. Further, in animal studies, most well conducted investigations report a lack of clastogenic effect in the somatic or germ cells of exposed animals (ICNIRP, 1998). Other investigations that require further attention relate to possible synergistic action of RF exposures with chemical or physical mutagens or carcinogens.

Most cancer studies of animals have sought evidence of changes in spontaneous or natural cancer rates, enhancement by known carcinogens, or alterations in growth of implanted tumours (ICNIRP, 1998). However, they have provided only equivocal evidence for changes in tumour incidence. Chronic RF field exposure of mice at 2-8 W/kg resulted in an SAR-dependent increase in the progression or development of spontaneous mammary or chemically induced skin tumours. In a further study, exposure at 4-5 W/kg, followed by application of a sub- carcinogenic dose of a chemical carcinogen to the skin, a procedure repeated daily, eventually resulted in a three-fold increase in skin tumours. However, at these high exposures, temperature-mediated effects cannot be excluded.

Studies in which cancer cells were injected into animals have reported a lack of effect of exposure to CW and pulsed RF fields on tumour progression. Progression of melanoma in mice was unaffected by daily exposure to pulsed or CW RF fields following subcutaneous implantation, and progression of brain tumours in rats was not affected by CW or pulsed RF fields following the injection of tumour cells into the brain.

Moderately lymphoma-prone Eµ-Pim1 oncogene-transgenic mice were exposed or sham-exposed to radiofrequency fields for 1 h/day for up to 18 months using pulse modulations similar to that used for digital mobile telephones. Exposure was associated with a statistically significant, 2.4-fold increase in the risk of developing lymphoma (Repacholi et al. (1997). This long-term study needs replication and extension to other exposure levels and animal models before it can

43 EMF Risk Perception and Communication be used for health-risk assessments. Further research is also needed to determine the significance of effects in this transgenic model for human health risk.

Although weak evidence exists, it fails to support an effect of RF exposure on mutagenesis or cancer initiation. There is scant evidence for a co-carcinogenic effect or an effect on tumour promotion or progression. However, only a few studies have been published and these are sufficiently indicative of an effect on carcinogenesis to merit further investigation.

Effects on other systems

Early signs of neurotoxicity are often behavioural rather than anatomical. While many studies have been conducted at high-levels of RF exposure few relevant studies have used low-levels. Some of the more important studies are described below.

The blood-brain barrier (BBB) is a specialised neurovascular complex that functions as a differential filter permitting selective passage of material from the blood into the brain. It maintains the physiological environment of the brain within certain limits that are essential for life. Although extensive previous research has been unable to reliably identify permeability changes at low levels of RF exposure, in recent studies, increased BBB permeability was reported for RF exposures at SARs as low as 0.016 W/kg. These studies need replication and extension to allow a better determination of any possible health consequence.

Exposure to very low levels of amplitude modulated RF fields were reported to alter electrical activity in the brain of cats and rabbits. These experiments need replication and extension.

Exposure to low-level pulsed and CW RF fields has been reported to affect brain neurochemistry in a manner consistent with responses to stress. Effects on behaviour and drug interaction have been obtained with the same exposure parameters. Replication studies are needed to establish and provide further information on these effects.

Pulsed radiation

Exposure to very intense pulsed RF fields suppresses the startle response and evokes body movements in conscious mice (ICNIRP, 1998). The mechanism for

44 EMF Risk Perception and Communication these effects is not well established, and is clearly associated with heating at higher absorbed energies.

People having normal hearing perceive pulse-modulated RF fields with carrier frequencies between about 200 MHz and 6.5 GHz; the so-called microwave hearing effect. The sound has been variously described as a buzzing, clicking, hissing or popping sound, depending on modulation characteristics. Prolonged or repeated exposure may be stressful.

Exposure to low levels of pulsed or CW RF fields may affect neurotransmitter metabolism and the concentration of receptors involved in stress and anxiety responses in different parts of the brain.

The retina, iris and corneal endothelium of the primate eye were reported to be susceptible to low-level RF fields, particularly when pulsed. Various degenerative changes in light sensitive cells in the retina, were reported at specific energies per pulse (10-µs pulses at 100 pps), as low as 2.6 mJ/kg after the application of a drug used in glaucoma treatment. However, these results could not be replicated for CW fields. Further replication studies are needed.

Epidemiological and human volunteer studies

Cancer disease

By far the greatest public concern has been that exposure to low-level RF fields may cause cancer. Of the epidemiological studies addressing possible links between RF exposure and excess risk of cancer, some positive findings were reported for leukaemia and brain tumours. Overall, the results are inconclusive and do not support the hypothesis that exposure to RF fields causes or influences cancer.

Review groups that evaluated possible links between RF exposure and excess risk of cancer have concluded that there is no consistent evidence of a carcinogenic hazard. In some studies there are significant difficulties in assessing disease incidence with respect to RF exposure and with potential confounding factors such as ELF and chemical exposure. Overall the epidemiological studies suffer from inadequate assessment of exposure and confounding, and poor methodology. Further studies are underway to evaluate potential carcinogenic effects of chronic exposure to low-level RF fields and more are needed.

45 EMF Risk Perception and Communication

Other outcomes

Other health outcomes investigated following RF exposure, include headaches, general malaise, short-term memory loss, nausea, changes in EEG and other central nervous system functions, and sleep disturbances. There have also been anecdotal reports from several countries of subjective disorders such as headaches associated with the use of mobile telephones. Whether exposure to RF fields at very low-levels can cause such subjective effects has not been substantiated from current evidence, but further research is indicated.

Individuals have claimed to be hypersensitive to electromagnetic fields. The most common symptoms are headaches, insomnia, tingling and rashes of the skin, difficulty in concentrating and dizziness. Given the limited evidence and widespread concerns that the above effects have provoked, more research is needed to determine if these health effects can be substantiated.

Adverse maternal health outcomes, particularly spontaneous abortions and haematological or chromosome changes, have been reported to occur in certain populations exposed to RF fields. Some of these changes have also been reported in users of video display units. Taken overall, the studies in this area have not substantiated these effects.

RF Field Research Needs

Following a thorough review of the biological effects from exposure to low- levels of RF fields, the following research needs were identified (Repacholi 1998).

Laboratory studies

1. Studies should focus on the potential for cancer promotion, co- promotion and progression, as well as possible synergistic effects. In this context, long-term studies of carcinogenicity in normal and sensitised or transgenic animals should be carried out to validate work reported previously and to determine the SAR thresholds for low-level RF exposure effects. End-points for these animal studies include genotoxic, immunological, and carcinogenic effects associated with chronic exposure to low levels of CW, pulsed, and modulated RF fields. The studies on the potential for DNA damage following low-level RF

46 EMF Risk Perception and Communication

exposure challenge the previous body of data and should be investigated further.

2. In view of the existing data on neurophysiological and behavioural responses to low-level RF exposures, animal studies are needed to determine the influence of exposures on central nervous system function, including melatonin synthesis, permeability of the BBB and reaction to neurotropic drugs.

3. Animal studies are needed to determine the influence of low-level pulsed RF fields on the structure and function of the eye. Only two laboratories have attempted this work and conflicting results have been obtained. If RF exposure affects any components of the eye, this would be an important area of public health concern.

Other research areas, such as teratology and effects on reproduction and development were deemed to be of a lower priority because there is very little evidence that such effects occur at low-levels of RF exposure.

Epidemiological studies

As a general principle, studies on populations exposed to high levels of RF are more likely to provide information regarding the existence of health effects, if there are any, even though such exposure levels may not be representative of general-population exposure.

1. Studies should be conducted on the possible relationship between use of hand-held mobile telephones and the incidence of:

• Brain tumours.

• Salivary gland tumours, acoustic neuromas and other head and neck tumours.

• Leukaemia and lymphomas.

2. Cohorts with high occupational exposure should be followed with respect to cancer. The highly exposed occupational groups have yet to

47 EMF Risk Perception and Communication

be identified completely, but are likely to include members of military forces, plastic heat sealers and welders, and physiotherapists. The identification of these groups would benefit from the development of individual RF dosimeters. At present there is no basis for restricting these studies to any specific type of cancer.

3. Adverse pregnancy outcomes in various highly exposed occupational groups should be followed in cohort studies.

4. Studies of risks of ocular pathologies in mobile telephone users and in highly exposed occupational groups should be considered.

5. Studies of populations with residential exposure from point sources, such as broadcasting transmitters or mobile telephone base stations present difficulties because of the low RF fields. However, in view of recent studies suggesting increased incidence of cancer in exposed populations, they should be investigated further.

Human volunteer research

Research needs include:

1. Symptoms reported by mobile telephone users and others exposed to RF fields, such as headaches, sleep disturbances and similar subjective effects, should be considered. EEG changes have also been reported to occur from RF exposure. It would be especially useful to appraise symptoms both prior and subsequent to changes in exposure. Their relation to RF exposure should be tested in laboratory (provocation) studies of sensitive individuals.

2. While some animal studies suggest possible RF exposure effects on the neuroendocrine, neurological and immune systems, it is not possible to extrapolate these results to humans. Therefore human volunteer experiments to test the physiological action of RF on these systems may provide further information related to adverse health outcome.

48 EMF Risk Perception and Communication

Further Reading

ICNIRP (1998) Guidelines on limits of exposure to time-varying electric, magnetic and electromagnetic fields (up to 300 GHz). International Commission on Non- Ionising Radiation Protection. Health Physics 74: 494-522.

NIEHS (1998) Assessment of health effects from exposure to power-line frequency electric and magnetic fields. Portier CJ and Wolfe MS (eds) NIEHS Working Group Report, National Institute of Environmental Health Sciences of the National Institute of Health, Research Triangle Park, NC, USA, pp 523.

NRC (1996): Possible health effects of exposure to residential electric and magnetic fields. National Research Council, Washington: National Academy Press

Repacholi M, Basten A, Gebski V, Noonan D, Finni J, Harris AW (1997): Lymphomas in Eµ-Pim1 transgenic mice exposed to pulsed 900 MHz electromagnetic fields. Rad. Res 147: 631–640.

Repacholi, MH (1998): Low-level exposure to radiofrequency electromagnetic fields: Health effects and research needs. Bioelectromagnetics 19: 1-19, 1998.

Repacholi MH and Greenebaum B (1998) Interaction of static and extremely low frequency electric and magnetic fields with living systems: health effects and research needs. Bioelectromagnetics (in press).

49

EMF Risk Perception and Communication

Improving EMF risk communication and management: the need for analysis and deliberation

Dr. Philip Gray Programme Group MUT, Research Centre Jülich, D-52425 Jülich, Germany Tel: +49-2461-613536, Fax. +49-2461-612950 E-mail: [email protected]

Abstract

Difficulties are often experienced between scientists, industry, regulators and members of the public in discussing the risks of electromagnetic fields. As in the case of other potential risk sources, one reason for communication problems is the existence of different understandings of risk. This paper discusses these differences as well as some current concepts for dealing with them positively within the framework of risk management. Risk is a wide-ranging phenomenon which can be examined from different angles. Natural scientists focus on a limited number of - in principle - objectively measurable dimensions, i.e. probability and certain consequences such as fatalities. Individual members of the public see risk in a multifaceted way, in which objective probability information plays a minor role, and causes and consequences are evaluated against multiple criteria reflecting personal and social context, including factors such as trust in operators and regulators. Both approaches are rational for particular purposes and situations, and both can play certain roles in the estimation and evaluation of risks. Recent treatments of risk management suggest new ways in which inputs to the process from a wide range of stakeholders can be conceptualised (e.g. Presidential/ Congressional Commission, 1997). The concepts of deliberation, analysis and their contribution to risk characterisation are especially valuable (NRC Committee, 1996). These concepts, and how they relate to scientific and lay views of risk, will be explained and a preliminary attempt made to apply them to EMF risk communication and management.

Keywords: risk concepts, risk communication, risk management, risk characterisation, deliberation, analysis.

1. DIFFERENT UNDERSTANDINGS OF RISK

Companies, regulators and scientists frequently experience difficulties talking to laypeople in connection with possible health risks of electromagnetic fields. One major source of problems are differing views or understandings of risks, as will be explored in this paper. The paper will go on to explore the general relevance of each way of looking at risk, and discuss how the interaction between

51 EMF Risk Perception and Communication the two approaches can be productively managed. First though, a general question will be considered: why is it important to try to understand lay people's risk perceptions?

Reasons for studying risk perceptions

As we shall see below, risk perceptions exhibit some fundamental differences from a scientific approach to risk. A scientist or administrator may therefore ask, why is it important to examine the lay view of risk? Is not the scientific approach not only sufficient but superior? There are certainly arguments in both directions on this issue (Pidgeon, 1998). Few would wish to deny that scientific knowledge is critical for weighing up risks on a societal level. However, there are strong practical and normative reasons for arguing that it is also important for those involved in risk management to study risk perceptions.

The normative arguments for risk perceptions having a role in risk management are first that, in a democracy, people ought to have an input into decisions that affect them. This does not mean that risk perceptions alone should determine decisions, but they should be considered. A further argument is that measurements of risk perception help to reveal basic values about risks and their management, such as an insistence upon equitable treatment.

The basic practical argument for giving risk perceptions a role in risk management is that it is important for engineers, scientists and decision-makers to understand risk perceptions because this can help them to understand the way the public reacts to different technologies. This improved understanding should lead to better communication and (in the best case) more trust between the different parties.

Risk perceptions can contribute directly to expert risk analyses in various ways. One way is that the 'social' risks involved in risk management may become clearer - for instance, in anticipating the societal reactions to given risk events. Possible human factors failures because of 'faulty' risk perception by plant operators may also be highlighted. More subtly, despite the differences between lay and expert approaches, some of the limitations of lay perceptions - cognitive biases and heuristics - can also influence expert assessments. Becoming aware of these effects can help to reduce errors. Finally, public perspectives on risks can in certain cases help to enrich expert analyses by challenging assumptions and providing local (or other special) knowledge.

52 EMF Risk Perception and Communication

Of course, risk perceptions are variable, often based on partial information, and have a different rationality than that of scientific risk assessments. Judging exactly how they should influence decisions is not easy and differs from case to case. Nevertheless, based on the above arguments it is concluded that it is important to study them and - in general terms - to find ways to take them into account in risk management decisions.

The nature of the scientific and lay approaches to risk respectively will now be described.

The scientific approach (risk assessment)

Natural scientists and lay people, as well as individual scientific disciplines, differ in the precise way they conceive risk. This difference will be referred to here as the difference between a 'scientific' way and an 'intuitive' way of looking at risk.

Risk is essentially an attempt to evaluate possible undesirable future states or consequences. Although it is related to measurable physical events, there is no single, universally valid or useful way to operationalise risk precisely. In this sense, risk is like other abstract concepts such as the health of the economy, for which no single indicator is sufficient. However on a general level, most descriptions agree that risk incorporates at least two dimensions: the uncertainty and the seriousness of the potential consequence.

Natural scientists aim to describe phenomena in ways that are generally valid and reproducible, avoiding personal evaluation. From this viewpoint, a risk can broadly be defined as a function of the probability - measured by as scientific means as possible - and the seriousness of the undesired consequences. Evaluation of the consequences is avoided by selecting certain easily quantifiable and clearly undesirable outcomes, such as deaths and serious injuries or illness. Since these consequences are, by definition, similar between different forms of risk, scientists tend to focus on measuring the probabilities of occurrence.

Various dimensions of probability may be selected, for instance the average individual probability of the undesired consequence, or the probability of groups of at least a given size suffering this consequence (group risk). Alternatively the probability and number of consequences may be multiplied to give an 'expected value' of harm in a given period. These are all different risk indicators, embodying different definitions of risk which are each useful for different purposes (Gray,

53 EMF Risk Perception and Communication

1996; Renn, 1992). None is sufficient in itself to describe the risk completely. In that sense, the choice of any given risk measure is not wholly 'objective', but a simplification which has to be justified with reference to its purpose.

The intuitive approach (risk perception)

When a layperson is asked to assess the size of a risk, they take a range of factors into account. These may include information about the probability of harm, but other factors are usually equally or more important. Some of these factors relate to the evaluation of the consequence, which, as mentioned above, the scientist tries to avoid (at least at this stage). These include the amount of dread, familiarity, latency of the effect, and the equity of risk distribution.

Other factors seem (among other things) to relate to the perceived likelihood of personally suffering the consequence: for instance, the perceived degree of control, and the amount of trust in the operator. Another important consideration seems to be the overall evaluation of the risk in comparison to the (personal) benefits.

In this intuitive approach, risk is a "multidimensional construct" which is based on a much broader range of factors than the scientific approach (Lichtenstein et al., 1978). This mass of different factors can be reduced statistically to one or two main groups, known as the dread factor and the unknown factor. The dread factor consists of characteristics such as uncontrollability, fear, involuntariness, and inequitable distribution. On average (across individuals and types of hazard) the dread factor is the most important determinant of the overall perceived risk from a given source.

It therefore seems that laypeople simultaneously assess their effective chances of being personally affected by the risk, and the importance of the risk based on evaluations of the consequences and of the reasons for being exposed. These evaluations may be both practical and/ or value-based.

Comparing the approaches

It might be thought that the intuitive approach is less rational than the scientific one. However, it actually faces different constraints and has different goals, so that it is misleading to compare it directly to the scientific approach: they are not true alternatives. The reasons for this are described below (see also Table 1):

54 EMF Risk Perception and Communication

• Risk perception is ultimately a way of dealing with the specific risks and benefits facing an individual (or societal risks from an individual viewpoint)5. Risk assessment seeks to be generally valid and rarely deals with individual circumstances or preferences.

• The urgency of decision is often greater in the case of risk perception.

• The consequences of an underestimate for an affected individual are much more serious than for a scientist.

• Risk perception depends on information of many types, received through multiple channels, only rarely including primary scientific data. Risk assessment deals mainly with scientific information, transmitted through defined channels.

• Risk perception is ultimately an individual process, although admittedly socially influenced. Risk assessment is the product of scientific teams and communities.

• Risk perception takes local, situation-specific information into account. Risk assessment generally deals in averages.

• The definition of 'benefit' varies: the individual may choose to value differently to the average, while the scientist takes the total societal benefit into account

• Risk perception is not a formal, explicit, nor even necessarily a conscious process; it is not clear that risk perceptions develop in orderly stages and even if stages can be differentiated, these are unlikely to be directly comparable to the stages of risk management (see below): it is more likely that risk perceptions have many types of origin.

5 The majority of risk perception studies in fact ask individuals about a wide range of risk sources, of which it is likely that many are not directly relevant to a given individual. Nevertheless because of the wording of the questions as well as the nature of the results, it is reasonable to assume that generally, individuals judge the risk sources from a personal or 'real life' point of view, albeit with varying levels of involvement.

55 EMF Risk Perception and Communication

Table 1: Factors affecting the relationship to risk of scientists and affected laypeople

ACTOR Personal Urgency Consequen- Access to Relation- Local Definitio Defin- relevance of ces of error scientific ship to knowledge n of risk ition of of risk solution (false inform- scientists benefit negative) ation Scientist Low Low Limited, High Member Low One- Societal career dimen- sional Affected High High Potentially Low to Consumer/ High(er) Multi- Personal layperson major, medium recipient dimen- (in risky personal sional situation)

Social Dimensions of Risk

Controversies over risks are not caused solely by differences in risk perception, or to put this another way, risk perceptions do not exist in isolation. They are both cause and product of many social factors. For reasons of space, only some of the main features and trends that are relevant to understanding concerns and controversies over EMF will be mentioned here.

First of all, it is perhaps obvious that, since lay people have little direct access to primary scientific information, such information is almost always mediated (transmitted) by other actors. This may be done by the mass media, authorities, independent scientists (usually in the service of some form of stakeholder), or industry. Because of this, the degree of trust placed in these actors is crucial. Trust is a reflection of the actor's previous record and relationship to the audience, as well as social conventions, status, personal preferences, and communication skills.

An important trend which sociologists have detected in (post-)industrial societies is towards increased individualisation and rising pressure on people to take responsibility for their own life and decisions. This is one of the elements of the so- called risk society, in which risks in general, and environmental risks in particular, have become central objects of attention (Beck, 1986). It is not only that there are (apparently) more sources of technical and technological risk, but also that people are more inclined to see technologies and actions in terms of risk and personal decision. This appears to be a fundamental trend which individual companies or governments can do little to influence directly. The positive aspect is that it implies the possibility of intelligent participation in societal decisions. This also suggests

56 EMF Risk Perception and Communication that more effort needs to be given than in the past to communication and dialogue about risks.

Such trends are also related to the perception of negative human impacts on the environment on local and global levels. This kind of perception is widespread, although the details vary between countries and regions. This currently forms part of the background against which any technological development is judged.

2. PERCEPTIONS OF EMF RISKS

• Studies of the risk perception of electromagnetic fields have been carried out using various methods, including psychometric techniques, mental models studies, and focus groups.

• The psychometric (questionnaire) studies indicate that people in general do not rate electromagnetic sources as being among the highest sources of risk. However, they distinguish between different kinds of EMF source. For instance, power lines are typically rated as presenting higher risks than other sources such as electric blankets and hair dryers (e.g. MacGregor, Slovic and Morgan, 1994). Furthermore, power line risks are rated as more 'known', more dreaded, having more severe consequences, being much less controllable and much less equitable than those from domestic electrical devices.

• Relatively few studies so far have looked at the perceptions of mobile phone risks. Although anecdotal evidence suggests that there may be more concern about these, a survey from Austria indicates that mobile phone risks are ranked relatively low in comparison to other hazards such as nuclear power stations or pesticides - although still over one- fifth of the sample described the radiation from mobile phones as "very" or "somewhat" hazardous (Weiss, 1998). A study of the risk perception of consumer products in Germany also found that products emitting EMF, including mobile phones, were rated as less hazardous than many other consumer products, e.g. household cleaners or genetically engineered foodstuffs (Schütz and Wiedemann, 1998). No psychometric study yet seems to have looked at the distinction between mobile phone use and living near a mobile phone relay mast.

57 EMF Risk Perception and Communication

• At first sight these findings do not seem to explain why the risks of EMF become controversial in certain real cases. However on closer examination there are some important clues, particularly in the differentiated perception of different sources. Most actual controversies appear to arise over the siting of relay masts and in relation to power lines. As mentioned above, power line risks are perceived quite differently from those of domestic electrical devices. Factors such as lower equity and lower control, as well as higher perceived risk, are indications that power lines belong to a different category of hazard, which might be described as large, infrastructural hazards. These hazards seem able to produce much more controversy than, say, domestic devices, because they differ on qualitative aspects (such as perceived equity) that are not covered by risk assessment.

• The relative low risk ratings for mobile phones appear consistent with this explanation, since the surveys quoted above did not ask about the relay masts, which appear to be much stronger sources of controversy in practice (e.g. McManus, 1998). However as mentioned above, no direct comparison of these two sources is available.

• Other studies suggest that laypeople's understanding of the physics of electromagnetic fields is generally somewhat limited, although not wildly inaccurate. Laypeople usually underestimate the rate of decline in field strength with distance from a source, and are unaware of the differences between electric and magnetic fields (Morgan et al., 1990). It would be wrong to suggest that limited knowledge is directly responsible for high risk perceptions, but in cases where someone is already concerned, it may be a contributory aspect.

Another factor which needs to be taken into account is that the studies mentioned so far deal in averages, and in practice there is a range of different perceptions of EMF risks. In fact, a minority of people in many countries see EMF as a much greater risk than the average, notably those who exhibit apparent 'electromagnetic hypersensitivity' (Berqvist, 1998). These people attribute various health symptoms or concerns to given EMF sources. Presumably beyond these relatively rare people there are many others who have no fixed position and who at least see EMF risks as plausible, if unproven.

58 EMF Risk Perception and Communication

3. FRAMEWORKS FOR RISK MANAGEMENT

Risk management seeks to identify hazards, characterise the associated risks, and take decisions about whether and how to control them (NRC, 1983). For many years this process has been defined under the dominant US model as a linear process in which considerations of values and other 'non-scientific' factors are kept clearly confined to the decision-making or risk management stage. This model, although useful on a technical level, fails to provide any orientation for dealing with the complex problems of risks over which experts and members of the public have different levels of concern. This section looks at how more recent models can help to provide this orientation.

One of the most important problems with the 'traditional' conceptualisation is that it fails to show the complexity of the characterisation process, including the fact that value judgements are involved in choosing how to measure risks, as implied by the discussion in the first section. In particular, the role of communication about risks is largely omitted. Furthermore, the linear depiction is misleading, since most risk assessment/ management processes are in fact iterative and involve feedback.

Frameworks

Two recent US frameworks (as well as certain European approaches) have begun to take some of these limitations into account. The first example is that of the National Research Council's (NRC) Committee on Risk Characterisation (1996) (Figure 1). First, the Committee describes risk characterisation as a "synthesis of information about a potentially hazardous situation" which meets the needs of the relevant parties, including decision-makers. However, the framework's key innovation is to describe risk characterisation as the interaction of parallel strands of analysis and deliberation. Analysis refers more or less to the methods traditionally considered to make up risk assessment: the application of scientific, engineering, legal and other forms of organised knowledge and formal methods to a problem. However, the NRC Committee (1996) argues in detail that analysis alone is insufficient to produce a useful and widely accepted risk characterisation; this requires deliberation too.

Deliberation means communication, discussion and debate about a problem, in order to frame the issue, increase understanding, and prepare the way for decisions. Each of the two aspects builds upon the other: "Deliberation frames analysis, and analysis informs deliberation" (NRC Committee, 1996: 20). The

59 EMF Risk Perception and Communication analytic-deliberative process clearly requires input from scientists and decision- makers, but also (as suggested by the above description) from other interested parties (stakeholders) depending on the type of problem involved. These may in principle include external consultants, citizens' groups, industry representatives or any other affected or interested party. The extent and type of involvement in the deliberative process required varies greatly according to the case.

Risk management and characterisation as an analytic-deliberative process

Learning and feedback

Public Officials Implementation Problem Process Selecting InformationSynthesis Evaluation Scientists formulation design options gathering Analysis Analysis Decision Deliberation Deliberation Interested/ affected parties

National Research Council Committee on Risk Characterization 1996)

Figure 1: A schematic representation of the risk decision process (NRC Committee, 1996: 28)

Another framework has been put forward recently by the Presidential/Congressional Commission on Risk Assessment and Risk Management (1997) (Figure 2). This sees risk management as a cycle (analogous to other management processes) consisting of problem identification, (risk) assessment, identification of options, decisions (option selection), actions (implementation) and evaluation. The last stage feeds back into a new round of hazard identification and/ or risk assessment, so this approach, like that of the NRC Committee (1996), moves beyond a simple linear picture. Even more importantly, stakeholder collaboration is placed at the centre of the cycle, linking all the other stages. Thus, this approach too makes the communicative or deliberative element a central part of the process.

60 EMF Risk Perception and Communication

Figure 2: The risk management framework of the Presidential/ Congressional Commission on Risk Assessment and Risk Management (1997).

As descriptions, these approaches are an acknowledgement of what actually happens in practice. Risk management is often not an orderly, linear process in which experts first produce a risk estimate upon which decision makers base their decision. Rather, hazards may be identified because of public concerns, and the terms of the risk assessment are framed by many implicit and explicit assumptions. The results of the risk assessment may be subject to controversy and revision (within and without the scientific community) before being finally accepted by decision makers, and the latter often engage in extensive consultations with other parties before finally making a decision. Even then their decision is often tentative and incremental, requiring review and further decisions in future years. Decisions often have to be taken before the full analytic process is even completed.

Relevance to differing views of risk

What do these new ways to conceptualise risk management suggest about how to deal with different approaches to considering risk, in particular the difference

61 EMF Risk Perception and Communication between risk assessments and risk perception? This issue will be discussed in general terms in this section, and subsequently applied to the issue of EMF.

The newer frameworks emphasise the involvement of various stakeholders, the role of communication between these stakeholders, and the complex interaction between various stages and aspects of risk assessment and decision-making. The NRC Committee (1996)'s approach goes even further in specifying that analysis is not in itself a sufficient process for producing a risk characterisation. Rather, the purpose and framing of the analysis are aspects that require deliberation.

The exact way in which deliberation is or should be achieved may vary greatly; for instance, it may or may not directly involve those affected by the risk. However, it is suggested here that in one form or another, one effect of deliberation will generally be to help to take people's risk perceptions into account.

This does not mean that people's risk perceptions drive the results of risk analysis. The ways in which risk perceptions may be relevant can be illustrated by drawing upon the list of arguments for studying risk perceptions, presented in the first section, summarised here under four headings.

(1.) In a democracy, people ought to have an input into decisions that affect them.

This argument suggests that people have a general right to be heard - and to receive information - in connection with developments that may affect them in any way, including in terms of risks. In some cases, normal representative democracy may suffice for this, but in others these traditional channels are not well adapted to dealing with specific local concerns and need to be supplemented by more direct methods of consultation or participation.

(2.) Risk perceptions reflect basic 'values' about risks and their management.

Risk perceptions are significantly influenced by concern about factors often omitted from conventional risk assessments, such as evaluation of the consequence, degree of personal control, and equity considerations. Several of these factors are quite legitimate or understandable dimensions, which could provide useful guidance on how to frame the risk assessment, i.e. what questions it should be designed to answer. Public values can also be a

62 EMF Risk Perception and Communication

valuable input to risk management decisions; however they should always be weighed up in the light of scientific analysis and other factors as well.

(3.) Laying the basis for good communication.

Entering into a dialogue with affected people or interested groups may take more time and effort initially, but it can help to improve mutual understanding and trust and hence provide a basis for good communication over the long term.

(4.) Enrichment of the decision.

A dialogue with affected people or interested groups can also lead to decisions that are better in various ways: by increasing understanding of the social dimensions of any decision, by providing additional e.g. local knowledge, and even by highlighting possible biases or hidden assumptions within expert assessments.

4. APPLICATION TO EMF

The discussion in the previous section is very general. In this section, a brief, preliminary attempt will be made to apply the frameworks described above to the issue of potential risks from power and radio frequency electromagnetic fields (EMF). The main messages of the frameworks can be summarised as:

• risk characterisation is an interactive process between technical analysis and deliberation;

• to achieve this, communication and dialogue are minimum requirements;

• involving various stakeholders at appropriate stages in risk characterisation can lead to better technical and social results.

How does the management of EMF risks measure up on these points?

Communication and openness

63 EMF Risk Perception and Communication

The actors responsible for managing possible EMF risks - particularly the electrical utilities - appear, on the whole, to have taken a positive and proactive approach to assessing and managing these risks over the last couple of decades. Within these efforts, the emphasis on open communication of information to the public has gradually increased (Kunsch, 1998).

On an international level there have been various steps to open up the overall scientific process in connection with EMF risks (Kunsch, 1998). The international EMF Risk Project has made various scientific meetings open to the public. Some industry-funded research bodies such as the Forschungsgemeinschaft Funk in Germany also have the specific goals of providing information and encouraging dialogue. The relevant regulatory bodies in various countries (e.g. Ireland, Sweden, UK) also seem to have been relatively active in providing information to the public.

Thus, without being able to make a detailed analysis here, the general impression is that the EMF community does place a relatively high value on communication and, to some extent, on allowing interested parties to see how the risk assessment process functions.

Encouraging broad-based deliberation

The more challenging aspect of the approaches presented here (especially the NRC Committee's, 1996) is to go beyond simply allowing the interested parties to know what scientists do, and to invite these parties to make inputs into the risk characterisation process. This implies that these stakeholders move from a passive to a much more active communication role.

At least one major EMF research programme does seem to have taken significant steps in this direction, namely the US's Electric and Magnetic Fields Research and Public Information Dissemination Program (EMF-RAPID), begun in 1992. Right from the start this programme has actively sought both experts' and laypeople's opinions in order to decide on its actions and on who should be included in the process (Portier and Wolfe, 1998). The programme's process of analysis and deliberation involves scientists, advocacy groups and regulatory agencies. Two-way communication with advocacy groups is achieved mainly through the national EMF advisory committee, on which these groups are represented. The committee also has a say in the membership of the EMF-RAPID toxicological review process. Some of the advocacy groups representatives also participate directly in the reviews, which

64 EMF Risk Perception and Communication seems to have positive effects (Portier and Wolfe, 1998: 300). Another productive means for two-way communication has been publication of findings on the World- Wide Web, with the opportunity for comment, which has been extensively used.

This programme seems to illustrate well certain concepts of the NRC Committee (1996) in action. Deliberation and analysis are clearly intertwined, and efforts have been made to broaden the range of input into the deliberative processes. Importantly, the results of this broadening are rated by the organisers as positive for the quality and the acceptability (to advocacy groups) of the process and its results.

5. CONCLUSIONS AND FUTURE STEPS

In light of the way that EMF risks are perceived, the brief analysis above suggests that the EMF risk management programmes discussed are moving in the right direction, at least according to the frameworks for risk management presented here.

This analysis, however, is essentially exploratory. In order to measure the risk management regimes properly against the proposed frameworks, more detailed and comprehensive work would be needed.

It is important to note that the frameworks do not provide any absolute standards. The required amount of participation in analytic-deliberative processes, and how this should be organised, can ultimately only be assessed on a case-by-case basis. One basic starting point for any given programme would be to carry out an evaluation of how its risk communication is seen by the 'users', whether scientists, regulatory bodies or laypeople (Gerrard, 1998).

In addition, there is much scope for taking the basic ideas of the NRC Committee (1996) as well as the Presidential/ Congressional Commission and adapting them to produce a more detailed general model of risk characterisation and management for EMF. Such a model might be able to provide more specific orientation than these general models as they stand.

REFERENCES

Beck, U. (1986). Risikogesellschaft. Auf dem Weg in eine andere Moderne. Suhrkamp, Frankfurt am Main. English edition: The Risk Society. London, Sage, 1992.

65 EMF Risk Perception and Communication

Bergqvist, U. (1998). Symptoms related to working with or proximity to electrical devices. In: R. Matthes, J.H. Bernhardt and M.H. Repacholi (eds.), Risk perception, risk communication and its application to EMF exposure. ICNIRP 5/98, Proceedings International Seminar on Risk Perception, Risk Communication and its application to EMF Exposure, Vienna, Austria, Oct. 22 and 23, 1997, 229-243.

Gerrard, S. (1998). Learning from experience: The need for systematic evaluation methods for risk perception and communication. In: R. Matthes, J.H. Bernhardt and M.H. Repacholi (eds.), Risk perception, risk communication and its application to EMF exposure. ICNIRP 5/98, Proceedings International Seminar on Risk Perception, Risk Communication and its application to EMF Exposure, Vienna, Austria, Oct. 22 and 23, 1997, 163-179.

Gray, P.C.R. (1996). Risk indicators: Types, criteria, effects. Julich: Forschungszentrum Julich GmbH, Studies in Risk Communication, 56. 105 pp.

Kunsch, B. (1998). Risk management in practice. In: R. Matthes, J.H. Bernhardt and M.H. Repacholi (eds.), Risk perception, risk communication and its application to EMF exposure. ICNIRP 5/98, Proceedings International Seminar on Risk Perception, Risk Communication and its application to EMF Exposure, Vienna, Austria, Oct. 22 and 23, 1997, 327-341.

Lichtenstein, S., Slovic, P., Fischhoff, B., and Combs, B. (1978). Judged frequency of lethal events. Journal of Experimental Psychology: Human Learning and Memory. 4, 551-578.

MacGregor, D.G., Slovic, P. and Morgan, M.G. (1994). Perception of risks from electromagnetic fields: A psychometric evaluation of a risk-communication approach. Risk Analysis 14 (5), 815-828.

McManus, T. (1998). Practical implementation of EMF risk communication. In: R. Matthes, J.H. Bernhardt and M.H. Repacholi (eds.), Risk perception, risk communication and its application to EMF exposure. ICNIRP 5/98, Proceedings International Seminar on Risk Perception, Risk Communication and its application to EMF Exposure, Vienna, Austria, Oct. 22 and 23, 1997, 315-326.

66 EMF Risk Perception and Communication

Morgan, M.G., Florig, H.K., Nair, I., Cortes, C., Marsh, K. and Pavlovsky, K. (1990). Lay understanding of low-frequency electric and magnetic fields. Bioelectromagnetics 11, 313-335.

NRC (National Research Council) (1983). Risk Assessment in the Federal Government: Managing the Process. National Academy Press, Washington, D.C.

NRC (National Research Council) Committee (1996). Understanding risk. Informing decisions in a democratic society. National Research Council Committee on Risk Characterisation. Washington, DC, National Academy Press.

Pidgeon, N. (1998). Risk assessment, risk values and the social science programme: Why we do need risk perception research. Reliability Engineering and System Safety, 59 (1): Special Issue: Risk Perception versus Risk Analysis, 5-16.

Portier, C. and Wolfe, M.S. (1998). Risk communication: The focus in the NIEHS RAPID program's review of EMF health hazards. In: R. Matthes, J.H. Bernhardt and M.H. Repacholi (eds.), Risk perception, risk communication and its application to EMF exposure. ICNIRP 5/98, Proceedings International Seminar on Risk Perception, Risk Communication and its application to EMF Exposure, Vienna, Austria, Oct. 22 and 23, 1997, 295-301.

Presidential/ Congressional Commission on Risk Assessment and Risk Management (1997). Framework for Environmental Health Risk Management. Final Report 1, Washington, DC.

Renn, O. (1992). Concepts of risk: a classification. In: S. Krimsky and D. Golding (eds.), Social Theories of Risk. Praeger Publishers, Westport CT, USA. 53-79.

Schütz, H. and Wiedemann, P.M. (1998). Judgements of personal and environmental risks of consumer products: Do they differ? Risk Analysis 18 (1), 119-129.

Weiss, B. (1998). Risk and public opinion. In: R. Matthes, J.H. Bernhardt and M.H. Repacholi (eds.), Risk perception, risk communication and its application to EMF exposure. ICNIRP 5/98, Proceedings International Seminar on Risk

67 EMF Risk Perception and Communication

Perception, Risk Communication and its application to EMF Exposure, Vienna, Austria, Oct. 22 and 23, 1997, 343-352.

68 EMF Risk Perception and Communication

EMF risk communication: Themes, challenges and potential remedies

Peter M. Wiedemann Programme Group Humans, Environment & Technology Research Centre Jülich D 52425 Jülich, Germany E-mail: [email protected]

Introduction

The EMF issue has played a role for quite some years now in the discussion about technical risks. Anxieties and fears are expressed again and again among the population, especially at sites for mobile radio transmitting stations. These fears are in part exaggerated. EMF is therefore also a communication problem, more precisely: a problem of risk communication.

In order to communicate better, it is important to deal with the following questions:

• What determines the difficulties in communication about EMF?

• How should communication about EMF be prepared?

• What has to be noted in communication about technical aspects of EMF?

• What is important in communication about risks?

• How should communication proceed in difficult situations?

• How should EMF communication be evaluated?

These questions will be addressed in the remaining of the paper. The goal of the paper is to improve the understanding of EMF as communicative issue and to give down-to-earth-guidelines for communication.

2. What Makes Communication Difficult?

69 EMF Risk Perception and Communication

Scientific risk assessments are essentially based on damage to be expected (fatalities, damage to health), i.e. on a more or less differentiated version of the elementary risk concept from insurance mathematics: risk = damage x probability6. Lay people, on the other hand, use a simpler and at the same time more extensive risk concept which, moreover, is dependent on the type of risk under consideration. Although the two aspects of 'damage' and 'probability' also play a role in the intuitive risk concept, other assessment criteria such as knowledge about a risk can also be of significance, depending on the type of risk source. Furthermore, essential risk evaluation dimensions for lay people include the catastrophe potential, voluntariness/involuntariness and controllability. Such aspects, however, are not taken into consideration by experts in risk evaluation. The resulting different risk evaluations then lead to a dilemma: only those arguments are recognised which correspond to one's own evaluation frame.

It is therefore evident, that any risk communication begins either with expectable or with factual differences in risk evaluation. Direct or instrumental goals of risk communication are then:

• avoiding conflict escalations in disputes about risks;

• minimising differences in evaluation or at least establishing a rational dissent on such differences.

Furthermore, risk communication also pursues a final goal. Risk communication as part of risk management aims at a fair distribution of risks and at an appropriate distribution of measures for risk reduction.

It should be noted, however, that differences about risks are unavoidable in a modern society. A societal consensus should be aimed at, but cannot really be achieved. The idea of conflict-free risk communication is as unrealistic as is the idea of zero risk. Whether and to what extent differences about risk evaluations lead to conflicts and whether these will escalate depends also essentially on how communication is performed.

The problem of risk communication in a mass and media society is communication itself:

6 In medicine, risk is defined as "the probability with which damage to health occurs in a population with a particular exposure" (Greim 1992, 5).

70 EMF Risk Perception and Communication

• More and more communication leads to an inflation of communication.

• More and more demands exist simultaneously and seek the public's attention on the opinion market.

• Communication must be louder and louder and more heated to be heard at all. It seems to be obvious: you are only noticed, if you exaggerate.

• In many cases, communication does not take place any more between people, but about them - in the media.

Communication is not a process with rules that only have to be observed in order for it to be successful. Communication is open, lively and full of surprises. Similar to road traffic, actual behaviour cannot be predicted from the rules. Minor changes may already have great effects: (1) How you address the other person, (2) What tone prevails, (3) How fast you come to the point, (4) How formally the situation is handled and (5) Where and when under what circumstances the discussion takes place.

Wrong Assumptions and Wrong Hopes

Communication is often based on wrong assumptions and false hopes:

• Facts speak for themselves. The thing is to correctly explain the facts and the communication partner will then also act in this sense.

• Statistical data and perfect logic are the best instruments to enforce the truth.

• "I know what you mean." - There is often the unrealistic conviction of having understood the other person correctly.

• "But what I am saying is quite clear." - There is a tendency to overestimate the clarity and persuasive power of one's own statements.

• "I know what impression I am making on my communication partner." - As a rule, the communication partner has often a different picture of us than we imagine.

71 EMF Risk Perception and Communication

Problems in Risk Communication

From a psychological perspective, risk communication is particularly difficult if those involved start from different risk concepts. The difference between lay people and experts plays a decisive role.

Conflicts between risk concepts

Expert evaluation Lay people's evaluation

• scientific methods • Intuitive methods • probabilistic concepts • no/yes • acceptability of the risk • safety • changing knowledge • is it or isn't it? • risk comparisons • discrete approach • average person • personal involvement

Figure 1: Problems in risk communication (after Powell & Leiss 1997)

With respect to EMF lay people show considerable gaps in know-how and knowledge. Nevertheless, they have an idea of how EMFs act and an opinion about the risk involved.

Thus, for example, no difference is made between high- and low-frequency electromagnetic fields and only a few prototype appliances and systems are mentioned as sources of EMF risks (mobile radio, high-voltage line, babyphone, clock radio).

Lay people basically evaluate risks of EMF with the aid of media information as well as general considerations and impressions. As a rule, they are not familiar with any scientific data/studies.

Individual cases dramatically described in the media play a particular role.

Lay people are often uncertain whether and to what extent EMF is a health risk. In this situation, they refer to "rules of thumb", analogies: "If the media report

72 EMF Risk Perception and Communication on EMF risks, there must be something in it" or "In other cases it had also been asserted that there would be no risks, ... and then how did it turn out ... ".

A particularly difficult group are those who feel damaged or impaired by EMF ('electrically sensitive' persons). Their risk assessment is fixed - no change is conceivable here.

Difficulties are also caused by the communication situation: discussions are often 'heated' and turbulent. This applies above all to group discussions. They require a different approach than individual discussions.

In the same way, the type of questions on EMF may cause problems. Difficult questions concerning the risk can be posed in the form of reproaches, accusations or attacks which then quickly lead to conflicts. This requires special communicative skills.

3. The Preparation of EMF Communication

The preparation of EMF communication is an important step. No discussion is free from surprise and sudden change - in principle, anything can become a case of conflict. This makes preparation even more necessary. Preparation concerns, on the one hand, the contents: you must know what you are talking about. On the other hand, the communication partner and the situation are of importance.

• What do I know about the communication partner(s) (interests, concerns, fears, attitudes)?

• What is my own attitude to the risk issue?

• What are the communication partners' expectations? And what do the communication partners know about EMF?

• Which questions/arguments do I have to expect?

• What strategy/argumentation should I pursue in my risk communication?

The better you are prepared, and the more you know, the better the discussion can be carried on.

73 EMF Risk Perception and Communication

Attitude and Atmosphere in a Discussion

Every discussion depends on the atmosphere. This means that it is important to create by one's own attitude a positive discussion atmosphere aimed at "understanding".

In spite of all possible differences, the important thing is respect for the other person. The communication partner must be unconditionally respected as a person. Another important aspect is fairness in a discussion. Everyone must be entitled to speak, criticise and respond to criticism.

Openness is a further important point. This involves (1) interest and goals, (2) planning and action as well as (3) data and other basic guidelines of importance for assessing the state of affairs.

Attitudes

Respect and Esteem Fairness Openness

The other person's perception is to be positively influenced by one's own attitude. He/she should:

• not feel attacked in the situation,

• be put in a positive mood, and

• be attentive to the arguments.

• Moreover, it is also important to ensure credibility.

Planning a Discussion

Apart from the proper attitude, it is important to plan a conversation. Preparedness is important.

74 EMF Risk Perception and Communication

Planning

Setting oneself realistic and reasonable goals Adjusting oneself to the situation Trying to control the discussion

Clarification of the goal concerns two problems: on the one hand, the question "What do I want?" and, on the other hand, the question "What can I achieve?". On the assumption that there can be no zero risk and no zero conflicts over risks, the following goals can be specified for risk communication:

• the reduction of risks according to the ALARA principle,

• the fair distribution of risks,

• the containment/mitigation of societal conflicts over risks,

• the turning of conflicts over risks back into a rational discourse and

• the legitimisation of risk-taking expectations.

The preparation for the situation is a second important element in planning. It involves the following questions:

• Who am I dealing with? What attitude, interests and positions will I encounter? What else is there at issue?

• What is the communication partner's attitude?

• What are the objections?

• What is his knowledge of EMF issues?

• How can the discussion proceed? Which disturbances, problems and difficulties could occur? And how can I overcome them?

75 EMF Risk Perception and Communication

The process of the discussion must also be planned. On the one hand, the communication partner must be given room to illustrate his standpoints. On the other hand, however, it is important not to lose control of the discussion. This means not to lose sight of one's own goals and also to explicitly pursue the issues of importance. One should not be forced onto the defensive.

This is of particular significance for group discussions. In addition to allocating the floor and integrating all participants, it is important to structure the discussion and illustrate the results.

Means for discussion control are:

• Specify the issue: "We are going to speak here about the possibilities of risk reduction."

• Specify one's procedure: "I would first like to describe the problem. We will then try to find out how we can improve protection."

• Point out digressions: "This is an interesting point, but at the moment we are still concerned with a different topic."

• Evaluate contributions by others with a view to the discussion goal: "What does this mean with respect to protective measures?"

• Summarise results in a discussion: "Let me summarise: We have achieved the following ..."

This type of communication for the control of discussions may be understood as summarising or commenting on what oneself and the other person does or should do in a discussion.

Maxims of Discussion Control

• Everyone has the right to advance his/her point of view.

• Everyone has the right to query or criticise the other person's point of view.

• Everyone has the right to respond to criticism.

76 EMF Risk Perception and Communication

Preparedness for Discussion - Some Useful Communication Techniques

In addition to the attitude to and planning of a discussion, it is also important to master and properly deploy discussion techniques. Of particular significance are: active listening, asking questions, arguing properly and coping with objections.

Discussion Techniques

Active Listening Asking Questions Arguing Coping with Objections

Active listening is an important instrument for recognising the communication partner's interests, intentions and objections. However, active listening does not only improve one's own information basis. It also contributes to the development of a positive discussion atmosphere.

As a rule, you concentrate on what you want to say yourself: which arguments could be of relevance and how to present them. This is necessary, but not sufficient. It is essential to adapt oneself to the communication partner, explore his information needs and listen to him. Active listening includes:

• attitude to the communication partner

• behaviour (proper and exact listening)

• signalling (illustrating what one has understood)

Particularly in the case of differences with respect to content, it is important to respect the communication partner as a person. And despite all differences, one should particularly note any points of agreement. In terms of behaviour, active listening means in the first place giving room to the communication partner. In addition, it means keeping eye contact and showing interest.

Signalling means that you express what you have understood ("Your important points are ..."; "Your particular concern is ..."). Anything not definitely

77 EMF Risk Perception and Communication understood should be clothed in a question ("Is it correct that you ...?"). And one should be open for corrections.

Two questions are essential in preparing for active listening: What are the assumptions with which the communication partner enters into the discussion? How will the communication partner interpret my arguments?

To begin with, all assumptions concerning the interests and attitudes with which the communication partner enters into the discussion should be listed. For each assumption, the statement should be put down which is important in this context - is my assumption correct or incorrect? Prepared in this way, you will better recognise critical points in a discussion and you can check whether your own assumptions concerning the communication partner are correct.

My assumption: "I assume What should be specifically that my communication noted in listening partner ..."

Is mistrustful Makes remarks on earlier negative experience with industry

Thinks that problems are Points to a lack of scientific played down studies

Furthermore, a list of one's own core messages should be drawn up. For each core message, it should be determined how this could be wrongly interpreted by the communication partner. And finally, consideration should be given to how the message would have to be formulated in order to prevent such misinterpretation.

With this preparation it is easier to discover potential misunderstandings and sensitive points in a discussion and counteract them.

78 EMF Risk Perception and Communication

Core message How could it be How would it have to be misinterpreted? formulated?

We comply with limit Anyone can say that. Compliance with the limit values. values is checked.

Our research has not After all, it has been Any research activity, including provided any evidence paid for by industry. our own, should be evaluated of hazards to man. according its scientific quality.

Asking Questions

Another important discussion technique is asking questions. If you ask questions, you will lead in a discussion, win time and be able to clarify misunderstandings. Asking questions is moreover a good approach for coping with objections.

Questions are meaningful in order to:

• clarify positions

• elaborate objections

• recognise reasons

• avoid confrontations

• improve the discussion atmosphere

Questions may be asked in different ways. Three types of questions are differentiated here.

The open question: How do you see ...?

The propositional question: What do you think of ...?

The question to cope with objections: What speaks against ...?

79 EMF Risk Perception and Communication

Questions should be simple. Whole series of questions are felt to be unreasonable or regarded as an interrogation. If you want to ask a series of questions, it is therefore important to announce this: "I would like to know your point of view more precisely. If you permit, I would therefore like to ask you some questions."

Argumentation

Anyone who wishes to convince needs arguments. Such arguments must be credible. Building blocks for good argumentation are:

• advancing appropriate arguments

• considering emotions

• using pictures

• illustrating conclusions

You cannot only convince with figures, data and facts. Examples and attractive models can also be used. You can appeal to reason, emotions and social bonds.

Arguments

Formal arguments: Requirements & Regulations

Scientific arguments: Figures, data and facts

Social arguments: Experts´ opinions

Social arguments: Social and emotional reinforcers

In argumentation, a three-step scheme has proved useful:

80 EMF Risk Perception and Communication

• Statement/recommendation for action: EMF risks are tolerable if the limit values are complied with.

• Arguments in favour:

− The limit values are derived on the basis of all available knowledge.

− The limit values consider the significant effects.

− The limit values contain an additional safety factor.

• Summary: These three facts corroborate ....

• Conclusion: that the safety margins exclude dangers to health.

4. Guidelines for EMF Technical Communication

Discussions on EMF are not only concerned with risks. Technical questions are also asked. Anyone who can answer these questions in a rapid and comprehensible manner will create an impression of competence in a risk discussion.

• How does mobile radio technology function?

• What power levels are used in mobile radio technology?

• What is the difference between thermal and athermal effects of EMF?

• Where do ELF and RF fields occur and how do they differ with respect to their effects?

Such questions should be answered in a brief and concise manner. It is important to: (1) Classify questions and problems (what is important to the communication partner?), (2) Explain in a comprehensible manner (simple language which is also understood by lay people); (3) Avoid giving a lecture - always ask whether you have been understood; (4) Provide assistance in understanding - translate scientific statements into everyday language.

81 EMF Risk Perception and Communication

You can prepare yourself for technical questions. And you can always keep information material on hand (pictures, figures) to illustrate your explanations.

5. Guidelines for Risk Communication

Risk issues are also addressed in discussions on EMF:

• How safe is mobile radio technology? Can mobile radio waves cause cancer?

• Are mobile radio waves dangerous for pregnant women?

• Can mobile radio waves influence brain waves or impair sleep?

• Are electronic devices such as PCs, alarm systems or cardiac pacemakers disturbed by mobile radio stations?

• Is the growth of plants/behaviour of animals disturbed by mobile radio waves?

Such questions can be answered with the aid of two different strategies. On the one hand, specialist arguments can be presented. This includes the explanation of limit values, comparisons, information about the state of the art and the quality of studies. If such knowledge is not available, argumentation is restricted to a credibility strategy.

Technical and scientific arguments

Argumentation with limit values, comparisons and information about the state of the art.

Referring to credible sources as argument

Argumentation with reference to scientific authorities and one's own person.

The Credibility Strategy in Risk Communication

82 EMF Risk Perception and Communication

• Illustrating one's own personal opinion

• Referring to experts

Example:

Due to the distance regulations which have to be complied with by the operators of systems, humans only come into contact with tolerable fields. This is also stated by the World Health Organization on which I base my arguments. And the WHO relies on the best scientists.

Company positive Company negative

Consistent corporate policy Frequent change in corporate policy

Competent management Weak management

Homogeneous interests within Conflicting interests within company company

Open for public matters and concerns Defensiveness and lack of sensitivity

Demonstrates social and Denies responsibility environmental responsibility

Renders account to the public Insulates itself

Subject to public control Internal control only

Positive experience with products Negative experience with products

Perceived benefit for public high Perceived benefit for public low

Perceived risks for public small Perceived risks for public high

Table 1: Factors influencing a company's credibility

83 EMF Risk Perception and Communication

Person positive Person negative

Personal aura Stereotyped behaviour

Competent appearance Unsureness becomes apparent

Shows specialist competence Shows himself (herself) to be uninformed

Sensitivity in handling public concerns Appears ignorant

Is perceived as honest, fair and upright Appears false, unfair and insincere

Demonstrates personal interests Hides personal interests

Clearly explains what he/she wants Remains unclear about his/her intentions

Also admits uncertainties Makes pretentious promises

Table 2: Features of a company spokesperson's credibility

Presentation positive Presentation negative

Informing in good time Informing too late and with delay

Keeping abreast of developments Providing obsolete information

Expressing oneself briefly and Presenting something at great length concisely

Referring to an audience's information Talking past the audience needs, value attitudes and ideas

Appealing to emotions Lecturing in an unimaginative and boring fashion

Making logical conclusions Conclusions remain unclear

Table 3: Factors influencing the credibility of presentations

84 EMF Risk Perception and Communication

Technical and Scientific Arguments in Risk Communication

Comparisons are important for understanding risk related information. Table 4 gives a survey.

Function Method

Comparison to illustrate the power level Comparison with a limit value; of a EMF emission source Comparison with EMF sources requiring risk-reducing measures

Comparison to illustrate the relative Comparison with other EMF sources significance of the EMF emission having the same impacts

Comparison to improve visualisation Giving examples of small probabilities (distances, quantities)

Table 4: Guidelines for comparisons

Comparisons can make risks and damage more comprehensible and can help to better explain the significance of a risk. In EMF comparisons, however, particular care should be taken: a suspicion of manipulation arises all too easily, and some comparisons also violate the risk perception of lay people. For example, it does not make sense to compare a technically induced additional risk with the risk of driving a car since, in the first case, an involuntary risk is involved and in the second a voluntary one.

Comparisons such as "risk of EMF versus risk of smoking" should therefore be avoided.

Case: Fear of radiation in the residential environment

• Basic situation: Person is afraid of electrosmog rays

• Procedure:

− the question is: When is the fear justified and when isn't it?

85 EMF Risk Perception and Communication

− illustrate distance values

− additionally: explain limit values

Even at a distance of only a few metres from the transmitting station there is no risk of an adverse effect on health due to the low field intensity.

A comparison with other EMF sources should be drawn carefully. It is best to compare installations known in the region (for example a television transmitter).

Case: Overestimation of mobile radio technology as an emission source

• Basic situation: Person regards mobile radio technology alone as the essential source of exposure

• Procedure:

− Compare power levels of base stations with broadcasting and transmitting stations

− Transmitter station approx. 50 W at highways

− Approx. 10 W in residential areas

− Television stations some 1000 kW

− The power level of a transmitter station is 16 - 80 times lower than that of a microwave oven.

The power levels of the mobile radio station should be compared with known, already accepted installations:

• The power level of all base stations is lower than the power of one TV parent transmitter.

86 EMF Risk Perception and Communication

• The power of the base station is of the same order of magnitude as that of a service radio.

Case: Comparison with natural background exposure

• The basic exposure due to natural fields is on average 0.003 W/m2

− After the final extension of the mobile radio networks, the load due to these fields will be 0.0002 W/m2

− Exposure due to mobile radio technology is thus clearly lower than the already existing average natural background exposure of 0.003 W/m2.

This comparison makes sense if the risk of mobile radio technology in general - in society - is being discussed.

The extension of the cellular telephone systems only causes a slight increase in basic exposure due to electromagnetic fields. The health risk from additional EMF exposure due to the extension of the digital mobile radio networks is thus only very small in comparison to current EMF exposure.

Limit Values

The argumentation with limit values requires specialist knowledge and some practice. Basically, this always involves safety issues. You must first draw the attention to "limit values".

"If we discuss possible risks here, it is important to consider the limit values to be complied with in the construction of plants."

"The question of whether or not there is a risk relates to the level of field intensities."

87 EMF Risk Perception and Communication

Why limit values?

• Everything might be turn into a danger.

• The important thing is "how much", that is quantity, concentration or intensity.

• People should therefore only come into contact with "such a quantity" of a substance that there is no damage

Limit values are protective values. This idea should be given priority. It should also be made clear that what matters is always the quantity/intensity/concentration - this accounts for the danger.

How do limit values function?

• Limit values function according to the safety principle.

• This means: a safety margin is added to the threshold value.

• Threshold value: that value which must be exceeded to obtain an effect.

• Safety margin: additional amount below the threshold value.

The threshold value is essential. It indicates whether there is still an effect (in the case of EMF - not a disease but feelings of nuisance). For reasons of safety, a safety margin is added to the threshold value in order to be completely on the safe side. This leads to the threshold value.

EMF limit values for base stations

• Limit value is expressed in the state of safety.

• Procedure:

88 EMF Risk Perception and Communication

− Starting from the highest possible transmitting power

− Consideration of distance values

− How far must you be from the transmitting station?

− 2 - 10 metres safety distance

The field strength and the power flux density increase very rapidly with the distance from the antenna.

In order to ensure human safety, safe distances have been introduced.

• At a typical mobile radio transmitting station (50 W), for example, nobody is allowed to stay longer than 6 minutes close (less than approx. 1.80 m) in front of the antenna in the direction of emission.

• In the area from approx. 1.80 to 4.50 metres persons may stay for 6 hours.

• Beyond a distance of 4.50 metres there is no restriction for staying. Anybody can live/work here and has nothing to be afraid of.

In the adjacent residential areas, actual values generally remain 100 to 1000 times below this WHO limit so that the more stringent limit values required by critical institutions are also complied with.

Sometimes, the difference in limit values is taken as an argument. The example then given is Russia where the limit is 1000 times lower. In this case, the implementation should be referred to which, in Russia, is not checked and enforced. Furthermore, it may be pointed out that the limit value requirements of environmental institutes and associations are complied with.

Different limit values?

• Proposals for limit values differ because different safety margins are

89 EMF Risk Perception and Communication

required

− WHO: 41.5 V/m = 4.5 metres distance from the antenna

− Critics: 20 - 8 V/m = 8 - 20 metres distance

• But: measurements show that the field strengths in the residential environment are generally 100 - 1000 times below the limit required by legislation.

• All limit value requirements are complied with.

6. Communication in Difficult Situations

EMF communication is sensitive. Everything can become a case of conflict here. "Quarrelling about" can be sensed in every topic.

Particularly critical moments are: (1) dealing with claims ("Prove that ... "); (2) dilemmas ("Where is the safety clearance? "), (3) reproaches, complaints, attacks ("What you say here is not true ... ") and (4) imputations ("You are withholding investigations").

Examples are:

• Where are the research results that rule out any hazard?

• Is mobile radio technology 100 % physically unharmful?

• After all, it is known today that EMF makes you ill.

• What you are doing is a long-term test with the population.

• Absolute safety was also always claimed for asbestos and wood preservatives. The same is now happening with electrosmog.

How difficult situations become evident

90 EMF Risk Perception and Communication

• Words are taken too seriously - "You are speaking here of subjective complaints - You seem to think that's just my imagination ... "

• Mistakes/slips of the tongue/failures are overinterpreted. - "Let's stick to what you said. This makes it clear for everybody how little attention you pay to the interests of the population ... "

• Rash interpretations - "This makes it quite evident again that your only intention is to placate."

• Critical interpretations - "You are always just trying to calm people down".

What is important in difficult situations?

Two things are important in difficult situations: (1) not every question must be answered in the form it has been asked. In particular, this also applies to the questioner's intention - this can also be reinterpreted. For example, you can turn an attack into a question or statement ("This topic is very important to you - I can see that.") Apart from such reinterpretations, illustrations are important. This means illustrating one's own intentions ... ("This is not what I meant, I would like to emphasise once more ... ").

Reinterpreting the other person's speech

• Returning to contents

• Taking no notice of attacks, reproaches

• Reducing positions to interests

• Reducing the question to a different one

Reinterpreting is important, if a question cannot be answered and if the question is intended to be unanswerable. For example, if the question relates to 100 % scientific safety. The answer can only be "no". It should therefore be

91 EMF Risk Perception and Communication reinterpreted: "You are interested to know how safe you are here (in the community, at the site etc.). I can give you the answer: "You are safe."

Illustrating one's own intentions

Saying:

• What you do, what you don't do

• What you (do not) want, what you are (not) driving at

• What you are (not) referring to

• What's it all about, what's up and what's not

• How something has to be evaluated and how it is not to be evaluated

It is often also necessary to illustrate one's own remarks to clear up misunderstandings or make something quite clear. Instead of "There is no risk" you can say "I assure you - and I do that with a clear conscience - there is no risk".

Illustrations can defuse conflicts, correct what has been said and prevent possible misunderstandings. In the following, you will find some examples demonstrating the application of these discussion techniques.

Dealing with claims and demands

This tower must not be sited here!

• Diagnosis: a position/claim is formulated here.

• Reinterpreting: asking questions.

• Reply: What are your reasons?

92 EMF Risk Perception and Communication

Claims should always be queried. They involve positions which always lead to a "yes" or "no", that is to approval or rejection. This should be circumvented by querying. Reasons can then be discussed later on again.

It is important to pass from a position/claim to the underlying interests and reasons.

Dealing with dilemmas

Dilemmas are situations in which you get stuck. They are characterised as follows: if you say "yes", this is classed as not being credible. If you say "no", this is taken as proof for the opposite side. Clear answers must therefore be given rapidly here. Otherwise you will lose.

Give me a guarantee that radio waves represent no danger at all for the environment!

• Diagnosis: dilemma, "no danger" cannot be scientifically proved.

• Reinterpreting: introduce the criterion of practical safety.

• Reply: according to practical criteria - which you and I assume in everyday life - I can assure you that there is no danger for you from this installation.

Attack on competence

Attacks can be directed against one's own competence (You don't understand a thing about it!) or against one's own credibility. Anyone who justifies himself here will lose and whoever makes a denial will heat up the conflict.

You are not a medical scientist!

• Diagnosis: attack on one's own competence.

• Reinterpreting: ask a question.

93 EMF Risk Perception and Communication

• Reply: that is correct. Do you think that only doctors are able to evaluate the situation?

That's industrial research!

• Diagnosis: attack on credibility

• Reinterpreting: question

• Reply: that's correct. Are you taking this opportunity to cast doubt upon the results? Do you have reasons?

After a Discussion

The evaluation of communication is an important task. You can only learn from experience if you evaluate. It is only by evaluating that you can realise what went wrong or right and what improvements are necessary.

• Evaluating the course of a discussion:

• What was positive?

• What was negative?

• Which questions remained open?

94 EMF Risk Perception and Communication

Risk Communications and the Management of EMF Risks

Gerry Kruk Gerry Kruk and Associates Communications Ltd 264 Parkside Crescent SE, Calgary, Alberta, Canada T2J 4J4 Tel: (403)225-1902, Fax: (403)278-9481 E Mail: [email protected]

INTRODUCTION: AN OVERVIEW OF RISK COMMUNICATIONS

The issue of public health risks from exposure to EMF is one of those controversies which will probably never be totally put to rest by scientific studies alone. However, controversies relating to the perceived EMF risks associated with proposed, new site-specific powerline or telecommunications transmission tower facilities can be effectively managed and resolved by the adoption of strategies based on the principles of risk communications.

In this paper, my objective is to advance this argument by doing the following:

1. Analyse public concerns about EMF from a risk communications perspective.

2. Identify the key risk communications strategies that contribute to consensual resolution of EMF risk controversies associated with site- specific EMF facilities strategy.

3. Explain the pivotal importance of public involvement as a risk communications strategy.

4. Profile a Canadian case study that illustrates the benefits of public involvement in resolving a non-EMF but similarly scientifically- complex and emotive risk controversy.

5. Discuss the role regulatory agencies can play in promoting public involvement to both developers & affected publics as a form of conflict- resolution in risk controversies.

95 EMF Risk Perception and Communication

1. EMF AS A RISK COMMUNICATIONS ISSUE

Why is the public so often concerned about facilities that produce EMF when scientific evidence overwhelmingly demonstrates that health risks from EMF are extremely low? Why is there often such a mutually frustrating non-dialogue between technical experts on the one hand and the concerned general public on the other hand concerning the highly emotive issue of risk? The answer to these questions is found in the way the public tends to think about risk issues, and how this approach differs from that of technical experts.

Technical Experts’ Approach to Risk

Put simply, the general public focuses on and attaches considerable importance to many different dimensions of risk than do technical experts. Risk experts tend to focus, somewhat clinically, on the quantifiable level of risk, which they recognise as a function not only of the consequences of a hazard but also its (im)probability. Technical experts are also aware that zero risk is impossible and tend to view reasonable risk-taking and technological innovation as important aspects of “social progress”. They are also very focused on what they view as society’s need for and benefits of science and specific technologies. Because of their technical training and responsibilities, they emphasise the importance of objective scientific facts in risk management decision-making and tend to instinctively view emotions and subjective perceptions as inappropriate considerations.

As a result, their priority is improving project designs and operational procedures to minimise risks and conducting quantitative risk assessments to confirm the effectiveness of their risk management initiatives. They understandably believe that when these scientific studies show that the risk associated with a project is or has been reduced to a low level, this project should be accepted by the public as “safe”. When the public refuses, they often tend to attribute this to either irrationality, ignorance or the self-serving tactics of uncompromising environmental groups, sensationalist media or opportunistic politicians.

The General Public’s Approach to Risk, including EMF Risks

In contrast, when the general public judges whether a project or activity is “safe”, they consider many other legitimate aspects of risk in addition to what science says about the quantifiable “level” of risk. For the public, the key issue is not the level but the acceptability of a risk; something is safe only if the risks are

96 EMF Risk Perception and Communication deemed acceptable. Even demonstrably low risks can be judged as unacceptable. Specifically, affected publics often react with alarm to EMF risks, however small, because of the following characteristics of EMF risks:

EMF risks are inherently frightening because they are said to pose a threat of particularly dreadful illnesses such as leukaemia and cancer, especially for the most vulnerable people such as young children, the unborn and the elderly. Anxiety is further increased because these threats are not brief and immediately apparent but are instead ongoing and delayed, even intergenerational. While power lines and transmission towers usually are disturbingly visible reminders of the ongoing risk, EMFs are invisible and not easily detectable. Sometimes the technologies involved are intimidating because they are so new, unfamiliar and even exotic. Because they are typically untrained in science, the public’s concern about EMF risk is further elevated whenever they are reminded of the uncertainty and disagreement amongst technical experts, believing that if even the experts don’t know whether power lines are safe, we should err on the side of safety. Public anxiety is further fuelled by dramatic and memorable media coverage of alleged EMF risks, with profiles of tragically sick children and interviews with worried parents and experts against a backdrop of intimidatingly large power lines.

Public objection to EMF risks relate also to fact that they are not only seen as potentially catastrophic because of the pervasiveness of the transmission facilities in urban areas but unfair risks as well. While all residents benefit from the electricity carried by power lines and all cell phone users benefit from the transmission towers, those people living near these facilities are the only ones to face the risks. In short, inherent in EMF risks is what some people view as a morally objectionable inequitable distribution of benefits and risks. Finally, because power line EMF risks are man-made, created by identifiable and enticingly deep-pocketed organizations whose motives and honesty one can question, the anxiety and anger of concerned publics tend not to be impressed by the fact that there may be higher risks from naturally occurring EMF. People will cut Mother Nature a lot of slack but not well paid corporate executives who seldom seem to live near their own facilities.

One can assume that a project proponent has diligently undertaken the technical work of risk assessment and mitigation to reduce EMF risks to the greatest extent reasonably achievable. Nevertheless, that risk, however low, may still be judged as unacceptable by the affected public if the proponent hasn’t also been forthcoming in sharing information and in providing meaningful opportunities for

97 EMF Risk Perception and Communication public involvement in the project planning process. In other words, the public tends to judge as unacceptable risks that they believe are being imposed on then involuntarily and which seem largely beyond their personal control. Similarly, they will also tend to reject even low risks if they are managed by people that seem disrespectful to the public’s feelings and approach to risk issues and managed by a process that is insensitive and unresponsive to the public’s specific concerns. And finally, the public usually responds negatively to risks created by organizations they feel they neither know or can trust.

The Objectives and Benefits of Effective Risk Communications

In short, high public concern and distrust are the defining characteristics of a risk communications challenge. When the legitimate, non-technical aspects of risk of concern to the public are not adequately addressed by the organizations responsible for the creation and regulation of the risk, the affected publics are likely to react in an angry and distrustful fashion -- or to use Peter Sandman’s concept, to feel “outrage” about risks, even when they are actually quite low. Risk communications requires not only the effective assessment and mitigation of the “hazard” but also the sensitive management of the public’s “outrage”. Affected publics must not only “be” safe; they also must “feel” secure; technical risk assessment and mitigation processes address the former while an effective risk communications program is necessary for ensuring the latter. Risk communications is much more about building trust and strengthening relationships than simply the clearer presentation of technical data. It is not that the public is angry because they are unable to understand the technical data; instead, one way that the public manifests its anger and distrust is through a disinterest in and rejection of the data. When the public distrusts an organization’s motives, they won’t even look at its data.

Therefore, the objective of risk communications is to facilitate the development of an informed consensus on “acceptable” risk and a co-operative approach to managing the risk in the desired fashion. Because it is based on awareness and sensitivity to the understandable causes of the public’s outrage, an effective risk communications strategy can produce the following practical benefits that are crucial to achieving this objective:

(a) Defuse the public’s risk-related emotions, such as anxiety, anger, even outrage, and

98 EMF Risk Perception and Communication

(b) Enhance public trust in the corporation or government agency in question, thereby

(c) Creating an atmosphere based on mutual understanding and respect in which there can be a more calm, candid & informed discussion of all relevant issues and data, including the actual, scientifically calculated level of “risk” (and regulatory safeguards) as well as the counter- balancing benefits of the proposed facilities. In this way, there can be

(d) A more co-operative/consultative assessment and management of the risks, thereby

(e) Helping the organizations acquire regulatory approval for project proposals in a way that is not only more timely and cost- effective but that also enhances its public credibility and strengthens its working relationships with key stakeholders.

2. KEY RISK COMMUNICATIONS STRATEGIES FOR DEFUSING RISK CONTROVERSIES

In view of the public’s approach to risk described above, organizations proposing new facilities that are likely to be viewed by affected residents as creating unacceptable EMF risks should adopt the following key risk communications strategies.

(A) Build Trust and Strengthen Relationships

When the key issue is the emotive one of risk, the principal concern of affected residents is their judgement of the reliability of a company to do “the right thing” to ensure the safety of their families, to do the right thing even when the public or the regulator is not looking over their shoulder. Trust is therefore pivotal to strengthening relationships, reducing public anxieties about risk and to resolving risk controversies. Investment in good-will from the outset pays large dividends in the long term; the converse is that a bad start will create a spirit of mistrust that can last for years. Negative first impressions tend to result in a quick and costly slide down a slippery slope into very adversarial, protracted and expensive regulatory battles that are destructive to long term relationships and public confidence.

99 EMF Risk Perception and Communication

To answer the question: “how do you build trust with the publics concerned about the safety of our facilities?” is to illustrate how many of the most important risk communications skills derive from our experience as private individuals in building and sustaining relationships rather than from our technical expertise as a “professional” risk manager. In short, trust is built by being:

• a good neighbour - caring, responsive, helpful, accessible;

• courteous - thoughtful and considerate, personable;

• open and honest - giving straight and complete answers; minimising jargon;

• admitting it when you don’t know (or are uncertain) - that, despite your expertise, you don’t have all the answers - you don’t even know what all the questions are;

• true to your commitments - living up to them, quickly and thoroughly;

• apologetic when you make a mistake - admit it, say you’re sorry, explain what you’re doing to ensure it doesn’t happen again, &, if need be, pay compensation

• respectful and empathetic - towards the public’s risk concerns, values & perspectives

• animated by a strong sense of social and moral responsibility - rather than constrained by a preoccupation with your organization’s “legal” rights and obligations.

(B) Share Project Information / Control Through Consultation

Clearly, risk controversies involve not only questions of technical substance but also “due process”, not only “what is to be done?” but also ”who decides?” Experience has shown that when people have a high level of concern about emotive issues like their family’s health and safety, when the public has little trust in business and government, and when people feel empowered by a sense of their own citizenship rights and political efficacy, they demand access to pertinent information

100 EMF Risk Perception and Communication and an opportunity to be involved in the planning of developments that are proposed for what they regard as their “backyard.”

Because risk issues involve personal feelings and values and not just “hard facts”, a project operator or proponent seldom can effectively address the public’s risk concerns by unilaterally commissioning yet another scientific study or by even installing more safety equipment. Instead, risk assessment and risk mitigation must be done consultatively, in collaboration with affected stakeholder groups. This is because earning trust and forging strong relationships with these publics is pivotal to calm analysis of the scientific data, to co-operative problem-solving and generally to the achievement of consensus on and public acceptance of the risks in question. Therefore, meaningful public involvement processes are at the core of any effective risk communications program.

Briefly, I believe the key characteristics of a meaningful public involvement program are:

(a) Try to avoid adversarial science; from the start, take a co-operative approach to the collection and interpretation of project-related data relating to risk assessment and environmental impact. Science alone, especially when conducted unilaterally by the project proponent, cannot resolve risk disputes because in the absence of trust, scientific findings tend not to be believed by concerned residents. Adversarial science confuses and alarms the concerned publics; in a low trust adversarial situation, science is used as a weapon rather than as a means of conflict resolution.

(b) Share project-related information early, continuously and candidly; if need be, review your organization’s traditional definition of “classified” information. In the absence of trust, the public will interpret negatively a perceived refusal to share information. When the general public is no longer angry and suspicious, their willingness and ability to absorb and work with (your) technical information is greatly enhanced. (To further reiterate what I said above: an inability to understand technical data is the consequence rather than the cause of the public’s outrage.)

(c) Consult the public even about the design of the consultative process - its structure, procedures, participants, issue-agenda, objectives and

101 EMF Risk Perception and Communication

schedule. This will help build trust by ensuring the residents feel confident this is a meaningful effort to obtain and respond to their input rather than a cynical exercise designed to co-opt or “sedate” them

(d) The public should control the consultative process; the principal focus of the process should be the issues of concern to the public. (Don’t prejudge what those issues are.) Representatives of the project proponent, the regulator and other government agencies should only serve as “resource people” to the residents on the PAC.

(e) Invite representatives of all key stakeholders, regardless of their project views, to participate in the consultative process; be flexible about admitting new participants to the Public Advisory Committee (PAC) even after the consultative process has started. Typically, the anger and distrust of residents participating in a consultative process continuously diminishes as they witness the project proponent’s openness, reasonableness and flexibility around the consultation table.(When the project proponent refuses to let opposing groups participate in a consultative process, it provides these “activists” with an opportunity to attack the motives and credibility of both the proponent and the consultative process. In contrast, inviting such critics to participate in a consultative process controlled by local residents often results in these same residents growing increasingly suspicious of the critics’ persistent unwillingness to engage in reasonable compromises.)

(f) Clarify what the “deliverable” will be and the role of the public in the final decision. The public should have the final say on the content of the “deliverable”, that is, the final report and recommendations.(This public report usually is submitted to the relevant regulatory agency as well as to the project proponent). To minimise misunderstandings and frustration at the end of the process, it must be made clear from the start that the proponent retains the right to decide whether to proceed with the application while the regulator retains the sole right to make the final decision regarding project approval and possible conditions.

(g) Articulate and Demonstrate Support for the Consultative Process. Senior representatives of the proponent & regulator should reinforce the credibility of the consultative process by personally reiterating how seriously they will consider the recommendations resulting from this

102 EMF Risk Perception and Communication

process. They also should tangibly demonstrate that commitment through co-operative behaviour at the PAC.

(h) Provide Project-Design Options Wherever Possible. To enhance the credibility of the consultative process, it should take place prior to the finalisation of any regulatory application and identify alternative design options that it can live with. And in evaluating design options, the proponent should also use “decision criteria” that not only reflect its traditional engineering & financial considerations but which also, wherever possible, respond to the legitimate concerns and aspirations of local communities.

(i) Create a non-threatening atmosphere at the consultation table so that all parties are encouraged to be candid, respectful and supportive of a co-operative approach to issues and a creative, mediation-based approach to resolving conflicts.

(j) Maintain ongoing communications with the grass-roots constituencies of the stakeholder representatives who are personally active in the public involvement process.

(k) The proponent also should commit itself to ongoing consultation and co-operative problem solving also during the construction and operating phase, should the application receive regulatory approval.

Case Study: The “NCACC” Public Consultation Process

In case this all sounds too good to be true, the following case study provides tangible and persuasive evidence of the efficacy of public involvement in helping to resolve risk controversies. This case involves a 1991 proposal by Canadian Occidental Petroleum (COP) to drill additional sour gas wells adjacent to major residential areas near NE Calgary. COP started gas production, pipelining and processing operations in the Crossfield reservoir in 1961. These operations were conducted safely for the following 30 plus years. However, city suburbs moved increasingly closer to the gas field. To minimise future land use conflicts, COP, in 1984, submitted to the Alberta Energy Resources Conservation Board (ERCB) a drilling application designed to expedite the depletion of the reservoir before the surface lands were needed for residential development. Some of the proposed new surface facilities would be approximately one mile from existing subdivisions.

103 EMF Risk Perception and Communication

Because of the emotional and technically complex nature of the issues, the drilling proposal was criticised and opposed by virtually all the dozen or so community associations representing the 140,000 residents of the affected part of NE Calgary, as well as by organizations representing the landowners in the adjacent part of the MD of Rockyview. Despite the fact that sour gas has never killed any member of the general public during the industry’s 50 year history in Alberta, the residents’ principal concern was public safety -- the possibility of an accidental release of poisonous hydrogen sulphide, and related public anxieties concerning the adequacy of setback distances and the effectiveness of emergency response plans. Other concerns included incompatible land uses, threats to property values, a perceived lack of fairness in the distribution of safety risks and economic benefits, and generally the “rights” of surface owners versus mineral owners and CanadianOxy.

NE Calgary Sour Gas Drilling Proposal, 1984-90: Traditional Approach

The sequence of events followed the usual scenario for risk controversies when project applicants follow the traditional “Decide, Announce, Defend” approach:

• In 1984, COP prepared and submitted its drilling applications, supported by all the technical evidence required by the ERCB, including the best risk assessments that available science permitted. And of course the application fully met all regulatory requirements.

• Nevertheless the application triggered an outraged response from community leaders. As community spokespersons saw it, it was totally unacceptable to permit a multinational oil company to further increase its profits by drilling for more poisonous sour gas when this gas wasn’t needed by Alberta consumers and when the memorable Lodgepole “blowout” of 1982 had, in their view, clearly shown how threatening & potentially catastrophic such operations were to local residents, who also felt they would get few local benefits.

• The media further stacked the decks against an informed and rational discussion of the safety issues by its potent combination of “human interest” sensationalism and infotainment

104 EMF Risk Perception and Communication

• In the face of the inevitable scientific uncertainties associated with risk assessments, there were demands for more studies, especially on plume dispersion in the NE Calgary area.

• As a result of this highly charged confrontational situation, regulatory hearings were repeatedly postponed and the ERCB applications were shelved in 1985 in view of the initiation by industry and the Board of the GasCon 2 Plume Dispersion and Risk Assessment Study. It was hoped this new “expert” scientific study would answer outstanding questions and ultimately reassure the concerned public on the safety issue.

• However, science alone, however reputable the scientists and compelling the factual information, seldom provides a magical silver- bullet resolution of emotional, value-laden risk controversies. In the absence of trust and effective forums for encouraging co-operative fact- finding and problem resolution, more scientific data simply fuels the adversarial process with competing data and experts. So it was in this NE Calgary controversy. When, in the fall of 1990, the ERCB convened a press conference to announce the reassuring findings of the GasCon 2 study, local community leaders expressed cynicism about the independence and reliability of what they dismissively called “your” study. It was clear that this detailed and field tested study, in the current adversarial environment, would not be taken seriously by, reassure or certainly not change the minds of drilling opponents.

Risk Communications Approach: Trust through Public Consultation 1991-2

Following the completion of the GasCon 2 study, COP decided to prepare new drilling applications. This time however, recognising the many legitimate non- technical aspects of risk that concerned the public and in response to both the letter and spirit of the ERCB’s 1989 Guidelines on Public Involvement, the company decided to undertake a major public involvement program. This would serve as a key component of a proactive corporate effort, in an environment of mutual respect and partnership, to work with local stakeholders for the purpose of co-operatively preparing a drilling plan and application that was based on a detailed appreciation and sensitive response to local concerns and feelings.

105 EMF Risk Perception and Communication

With the assistance of a senior member of the ERCB, the eighteen concerned community associations agreed to send representatives to participate in a multi-stakeholder committee with representatives from CanadianOxy, the Board, the City of Calgary and the MD of Rockyview, the Calgary Board of Health, the Calgary Fire Dept. and Alberta Environment. The committee came to be called the Northeast Calgary Application Consultation Committee (NCACC). Its agreed-upon purpose would be, in a calm, non-threatening, environment characterised by mutual respect, candour and openness, to identify and address all the issues and questions of concern to the public, to review CanadianOxy’s preliminary project proposal, and if possible, to co-operatively develop a consensus position on the issues. The final deliverable would be a written report to the ERCB on the NCACC’s deliberations and their results, including recommendations that enjoyed strong (although not necessarily unanimous) support amongst the community participants on the Committee.

The residents decided that eight months would be adequate to complete this mandate of the NCACC. They succeeded in meeting this self-imposed completion date by submitting their report to the Board on June 30, 1992. During this consultative process, the community representatives developed a general consensus that the proposed new wells could be drilled and operated in a safe manner, without creating undue land use conflicts and with significant economic benefits to the Province and local economy. The bottom line was that the residents, in their NCACC Report, recommended to the ERCB that:

“On the condition that certain stringent recommendations are accepted and implemented... additional sour gas developments in the NE Calgary study area could proceed.”

Earning Informed Public Consent through Public Involvement

Through this relatively expeditious NCACC process of candid dialogue, co- operative fact-finding and good-faith bargaining, the community representatives and CanadianOxy steadily built up mutual trust and forged compromises that largely resolved the emotional issues that had long troubled their relationship and delayed regulatory approvals. The NCACC process was able to accomplish this because of the way the interpersonal dynamic associated with meaningful consultation can assist the proponent of a controversial project to effectively address all of the non- technical dimensions of risk that are of concern to the public and consequently have a transformative effect on public attitudes.

106 EMF Risk Perception and Communication

The NCACC process was extraordinarily successful in defusing public anxiety, clearing the way to a more calm and informed discussion of all the issues, and ultimately, in earning public support for CanadianOxy’s sour gas drilling proposal. It achieved this outcome by effectively addressing the risk concerns of the public in the following ways.

Share Project Control: The NCACC responded to the public’s legitimate demand to have a meaningful opportunity to be involved in the pre-application project planning process.

• There was no effort to exclude any stakeholder group despite their views on the project. The NCACC process was clearly controlled by the residents; COP and ERCB reps were simply “resource” people, answering residents’ technical and process questions.

• An impartial facilitator helped to assure the public representatives on the NCACC that, despite their relative unfamiliarity with the technical issues, they would not be at any disadvantage in the Committee’s discussions on the identification, clarification and resolution of the impact mitigation and risk management issues of concern to them.

• The public input was meaningful. Instead of simply requesting community feedback on a completed drilling plan, CanadianOxy indicated it would not prepare a formal well licence application until the community representatives provided their input on a “preliminary project description”. The communities were offered project design options, as well as an open, accommodating process for evaluating and supplementing those options.

• To emphasise the serious consideration that CanadianOxy and the ERCB would give to the residents’ concerns, and to ensure transparency and accountability for the NCACC process, detailed minutes were kept of all meetings, recommendations and decisions.

• All requested information was provided, and on the few occasions when the residents requested it, funding was provided for the services of mutually acceptable independent experts to provide additional information on particular issues. In short, instead of adversarial science, there was a collaborative approach to fact-finding.

107 EMF Risk Perception and Communication

• The NCACC process also illustrated how the public’s failure to understand technically complex risk data is typically a “result” rather than a “cause” of public outrage. For example, as their previous anger and suspicion dissipated during the consultative process, the community leaders on the NCACC were increasingly prepared to take at face value the scientific findings of the GasCon 2 study that many of them previously tended to dismiss as the biased work of less than impartial industry and regulatory organizations.

Build Trust and Credibility: In order to be viewed as a trustworthy individual or organization on the emotive subject of risk, it is not enough to be seen as technically competent, which the public usually just takes for granted. As in personal relationships, we also have to be judged as open and honest, as well as dedicated and committed to the issues of concern. However, the attribute which overwhelmingly determines how trustworthy you are seen on a risk issue is how empathetic you are to the public’s concerns, feelings and perceptions. People don’t care what you know until they know that you care.

Consultation processes are ideal opportunities for organizations to reassure the public, by both word and deed, that they genuinely respect and empathise with the public’s risk concerns and perceptions. So it was for CanadianOxy and the ERCB on the NCACC. They acknowledged the legitimacy and reasonableness of the public’s non-technical risk concerns. They took care not to alienate concerned residents by suggesting that they saw their role as educating the “uninformed, irrational and therefore needlessly frightened public.” Instead, CanadianOxy and the Board made it clear that their objective was to understand and effectively integrate the residents’ concerns and values into the project’s design. Rapport and trust were also earned by a willingness to candidly acknowledge past shortcomings and current uncertainties. Above all, no effort was made to pressure the residents on the question of “acceptable risk”; instead, they were simply provided the information necessary for making up their own minds. In this non-threatening, respectful environment, it was possible to use a wide variety of risk comparisons to provide perspective.

Fairness: After the issues of “control” and “trust”, probably the next most important non-technical dimension of risk that can enrage the public is the feeling that their community is being unfairly singled out to bear, to an inordinate degree, the risks associated with industrial operations like sour gas that benefit all the residents of the province. The NCACC process permitted this fairness issue to be defused because of the opportunities it provided to explain a number of relevant

108 EMF Risk Perception and Communication points: that many other communities in Alberta have sour gas operations adjacent to them; that there are in fact major counter-balancing local benefits for the Calgary area, and that CanadianOxy recognises the legitimacy of the fairness issue and seeks to address it through its hiring, procurement and donations programs. Because of the trust that had already been built up and the reassurances provided on public safety, the fairness issue could be addressed without the appearance of bribery. Public safety was the priority issue but fairness was treated as a legitimate separate issue.

Fear of a Potential Catastrophe: Clearly, in view of the Lodgepole experience, it was not unreasonable for the public to fear a potentially catastrophic accidental release of poisonous hydrogen sulphide. However, the calm and rational atmosphere of the NCACC provided the company and Board with the opportunity to explain the physical characteristics of the Crossfield reservoir and of the local topography and meteorology that together greatly reduced both the chances and consequences of any accidental gas release.

This atmosphere also made it possible to discuss the major review conducted by the Board and industry of the “memorable” Lodgepole incident as well as of sour gas regulations generally. Information on the subsequent further upgrading of regulatory safeguards and the consequent enhanced performance of the sour gas industry greatly reassured the public. This dialogue also helped to transform the Lodgepole incident from an alleged sign of the riskiness of the sour gas industry to a symbol of the industry & regulatory diligence. Such a constructive dialogue was not possible during the previous adversarial review process.

The Benefits of Public Involvement

Besides adding real value to project planning, public involvement brings many benefits:

(a) By acknowledging the legitimacy of the public’s right to be meaningfully consulted on projects that may affect them, public involvement processes help to defuse emotions.

(b) The co-operative interpersonal relationships that characterise and lubricate public involvement processes provide an ideal opportunity for empathetic risk communications, that is, for demonstrating your organization’s understanding of and responsiveness to the public’s risk- related concerns. Such a caring attitude enhances an organization’s

109 EMF Risk Perception and Communication

credibility on emotive risk issues and reduces the sense of adversarialism and confrontation, further defusing the emotions that act as a barrier to constructive dialogue and problem-solving.

(c) Displaying empathy and sharing control also earn an organization increased public trust and strengthen its working relationship with local communities.

In short, the NCACC experience demonstrated that this type of multi- stakeholder consultation process has unique value. It combines the scientific rigour of environmental impact assessments and the economic rationality of cost-benefit analysis with the procedural fairness of quasi-judicial regulatory hearings and the democratic principles of public participation. In so doing, such public involvement strategies are pivotally important to the process of defusing the public outrage and resolving risk controversies in a way that permits the public to acquire a constructive understanding of the relevant scientific data and ultimately give informed consent to development proposals.

(C). Meet the Information Needs of the Mass Media

Risk communications involves respectful listening and meaningful responsiveness to the public’s perspective on risk issues. However, organizations involved in EMF risks also have important messages to communicate. They therefore must keep in mind that, far from being an alternative to forcefully advocating the benefits and safety of their operations, the adoption of a risk communications and public involvement strategy actually prepares the way to a more credible and focused presentation of your case in the court of public opinion and a more receptive public response to your message. Such advocacy necessarily involves the use of the media as part of the organization’s legitimate effort to communicate its (EMF) risk messages to the general public. Effective media relations is an important component of any risk communications strategy.

It is from the mass media that citizens today obtain most of their information on “risk” issues. Most of the political pressure on governments to “do something”, to “protect” public safety by reducing or avoiding these risks ultimately derives from what the public sees and hears in the media. By understanding the biases, motivations, and needs of journalists, we can more effectively communicate essential risk messages to the public through the media. In the final analysis,

110 EMF Risk Perception and Communication proactively improving the media relations performance of news makers is more productive than blaming journalists for being journalists.

Institutional “Driving Forces” Behind Media Approach to Risk Reporting

There are a number of factors or “driving forces” that fundamentally influence or bias the media to cover risk issues in a largely predictable way. These are systemic factors that risk communicators must keep in mind as they endeavour to communicate through the media.

News as “Infotainment”. The media is a “business” in a competitive market-place. As such, broadcasters are concerned about their ratings and publishers about their circulation. Media outlets feel they must compete not only with each other but with the entertainment business. Several consequences result from the media’s concern that the news be entertaining. Firstly the mass media feels compelled to adopt a journalistic value system that defines “news” as revolving around crisis, conflict and controversy. Accordingly, the media is usually more interested in drama, sensationalism and political tactics rather than the technical substance or socio-economic consequences of public policy issues. Secondly, partly because of this preoccupation with conflict and controversy, the mass media tends to define “objectivity” not as a search for the empirical “truth” but rather as simply “balance”. They define this “balance” as the provision of “fair” coverage of opposing viewpoints. Little regard is given for how reasonable or supported by scientific proof or professional peer opinion that opposing viewpoint is. And thirdly, media coverage tends to be stimulated by or revolve around some kind of a news “hook”-- usually a spectacular, photogenic event. The result is an approach to news that the media itself describes as “if it bleeds, it leads!” What has been called the “dumbing down” of the news has been further accelerated by the increased competitiveness triggered by the advent of 24 hour news programs.

Careerism. Reporters’ career aspirations tend to further reinforce the institutional bias in favour of sensationalist and entertaining “bad” news. To paraphrase the esteemed journalist Theodore White: You don’t make your reputation as a reporter by praising anybody. You do it by gouging a chunk of raw and bleeding flesh from the system.

Untrained in Science. Most of the reporters who cover risk-related incidents or issues are not well trained in science and are usually not very interested

111 EMF Risk Perception and Communication in having you take them through an academic exercise. Consequently, journalists usually find it easier and professionally less risky for them to focus on the “politics” than the science of risk issues.

Self-image as Public Guardians: Especially since the Watergate scandal, many journalists have tended to exhibit strong distrust in “the powers that be” and to define their mission as protecting the little man from the perceived machinations of big government, big business and big unions. As one of my journalist friends put it: “Our role is to be a watch dog for the underdogs, not a guard dog for the top dogs”. In terms of risk reporting, this orientation often leads to excessive negativism and cynicism towards the motives and truthfulness of corporate proponents and government regulators.

Time and Space Constraints. Finally , the media’s coverage of risk issues is affected also by the time and space constraints and short deadlines faced by the media. This constraint has been made even more significant by the endless nature of today’s news cycle. Of course, accuracy is sacrificed at the altar of scooping the competition and of style - it’s a lot easier to write eloquent prose when you’re not grappling with all the data, qualifications and complexities originally communicated by the expert.

Risk Perceptions. Journalists, like the general public, are non-technical people, and therefore tend to give considerable priority, when judging the acceptability of a risk, to the non-technical social, ethical and due process dimensions of risk.

Media Treatment of EMF and Other High-Tech Risk Issues

How do these “driving forces” influence how the media tend to cover risk issues?

1. Because the media seek to make their reporting as “interesting” as other forms of entertainment, journalists typically show relatively little interest in complex scientific or public policy issues. The only exception tends to be when the issue has been made “urgent and exciting” and of “human interest” because of some dramatic “study”, accusation, or incident. The media is more interested in “juicy stories” than in “dry facts”.

112 EMF Risk Perception and Communication

2. The media’s preoccupation with crisis, controversy and dramatic news “hooks” often leads journalists to show greater interest in apparent technological failures than successes, in sensationalist “bad” news -- catastrophic events that can be described using graphic language and memorable “high-impact” visuals. Fear, especially of exotic, unfamiliar or omnipresent technologies, is a very marketable media theme -- the red meat they seek.

3. A number of consequences often flow from journalists’ unfamiliarity with science. Firstly, many important omissions often characterise their reporting on human risk issues. For example, journalists often miss the significance of exposure levels and dose; also, the media rarely mention surrounding naturally-occurring risks. As well, in much media coverage, risk is not “assessed” but simply “assumed” to result from the mere presence of a particular man-made chemical or effect. Health symptoms are reported, but issues like causal relationships and the actual degree of risk are seldom investigated. This often results from the media’s uncritical reporting of human-interest anecdotal “evidence”, such as informal health surveys by residents which describe their various ailments & their “belief” that these ailments were caused by some nearby industrial facility.

4. As well, because of this media and public tendency to simply view something as either “safe” or “risky”, balance and perspective, such as can be provided by comparing a technological risk with a more common risk, such as crossing the street, is usually absent.

5. In policy areas as complex and potentially emotive as public safety, the media’s approach to “objectivity” often leads them to trivialise science by reducing it to a conflict of sound bites from competing authorities. Because the media feels compelled, even when there’s an overwhelming scientific consensus, to seek out contrary points of view, some-times calling on media-wise spokespersons of doubtful technical expertise and repeatedly referring to discredited studies, readers & viewers are too often left confused.

Unsurprisingly, in view of this media role definition, environmental activists tend to supply the ideas, assumptions and arguments that form the less than impartial context for most media reports on environmental risk issues. For example, a 1993

113 EMF Risk Perception and Communication survey of 512 American journalists found that, of those sources for story ideas and data that are most likely to be biased -- environmental / consumer advocates and business, the former are relied upon by the media 8 times more often than the latter.[14] These patterns of risk coverage by the media tend to distort the science of EMF risk, exaggerate the actual risks of EMF technologies, and obstruct the industry’s effort to obtain impartial coverage of credible scientific studies and their own risk management activities. They probably also intensify public uncertainty, fear and opposition to existing and proposed EMF facilities.

Media Relations Strategies Concerning EMF Risk Issues

In view of the above, the following are some recommendations on media relations strategies to assist in communicating to (and through)the media on EMF risk issues.

(a) Proactively build relationships, trust and understanding with the media. Build up a reserve of goodwill which can be called upon in controversial situations.

(b) Demonstrate your sensitivity to the due process and ethical dimensions of risk.

(c) Use more personal approaches to inform journalists (i.e. field tours, interviews).

(d) React promptly to information requests; be helpful, accessible, open and honest Ensure the media have quick access to a credible, trusted, scientific sources. so that there is no excuse for the media to not incorporate your views / response in any article. Negative statements by the media about your organization and technology tend to quickly become seen as facts by concerned publics if you do not quickly correct or challenge them.

(e) In interviews or releases, simplify your message into a few key points which can be supported by evidence in the form of not just key statistics but illustrative risk metaphors and comparisons, memorable anecdotes, and credible third party endorsements as well. (For example, another benefit of effective public consultation is that previously concerned but

114 EMF Risk Perception and Communication

now reassured residents can become credible media spokespersons on risk issues.)

(f) Provide information that puts risk into perspective. Distribute fact sheets on the most important risk issues associated with your project as well as a detailed set of Q’s and A’s on the most frequently posed questions. These practical tools provide your people with a shared framework, data base and inventory of useful risk facts and comparisons. Provide journalists with a listing of the types of questions they could use to probe the accuracy and objectivity of the advocates of studies which allege a serious EMF risk to public safety.

(g) Provide your risk information as an interesting, socially relevant “story” (the “hook”). Translate Risk Assessments into practical “so what” statements for local audiences.

(h) Develop a clear, quotable way of stating the scientist’s usual boring, fence-sitting position (i.e. “ on the one hand, on the other hand”; “we need to do more research”).

(i) Avoid jargon, clarify technical terms, test for understanding.

(j) Put the risks of your proposal in perspective by comparing them to the alternative risks -- the risks of doing nothing and / or of alternative strategies.

(k) We cannot expect the media to function as industry’s “spear-carrier”. However, at the other extreme, if you can document how your organization has been the victim of unprofessional conduct by a media institution, you can often address this situation by holding that media accountable for violating their own journalistic standards.

Media outlets care about their professional credibility as do individual journalists. There- fore, they cannot afford to be indifferent to well founded complaints about their violation of their own code of conduct. Therefore, such exercises in scrutiny and accountability tend to contribute to greater accuracy, fairness and objectivity in reporting. For example, the CBC states that its information programs “must reflect (the following) journalistic principles:

115 EMF Risk Perception and Communication

Accuracy The information conforms with reality and is not in any way misleading or false. This demands not only careful and thorough research but a disciplined use of language and production techniques, including visuals.

Integrity The information is truthful, not distorted to justify a conclusion. Broadcasters do not take advantage of their power to present a personal bias.

Fairness The information reports or reflects equitably the relevant facts and significant points of view; it deals fairly and ethically with persons, institutions, issues and events.

Application of these principles will achieve the optimum objectivity and balance that must characterise CBC’s information programs.”

Clearly, journalism that consistently reflects such principles will contribute to the calm and informed discussion and rational resolution of risk issues. It will contribute to this goal by providing the public with the accurate, understandable, objective and complete information from the media that it needs and deserves, & which the media is obliged to provide.

The Alberta EUB’s Public Involvement Guidelines

As I have tried to emphasise and demonstrate, effective public involvement strategies are a key aspect of risk communications and are pivotally important to the effective management of risk controversies. However, it is sometimes the case that, from the start, corporations or government agencies promoting controversial projects are sceptical or simply unaware of the benefits of public consultation, while the concerned/affected publics may fear consultation as a recipe for co-option. And later on, because active risk controversies result in such high levels of polarisation, distrust and mutual animosity, neither the project proponent nor concerned publics are in the mood for belated consultation programs.

However, this reluctance and skepticism can be overcome by proactive promotion of public consultation by the regulatory agencies who must approve project applications and if necessary, adjudicate the applications through adversarial, quasi-judicial public hearings. A very successful example is provided by the Public Involvement Guidelines (Information Letter 89-4) issued in 1989 by the Energy Resources Conservation Board which must review and approve all energy related project applications & operations in Alberta, Canada.

116 EMF Risk Perception and Communication

In its Guidelines, the Board clearly articulated its expectation that project applicants must offer affected residents meaningful opportunities to become involved in project planning. In short, as the Board indicated in its Guidelines, “those Albertans who are -- or who perceive themselves to be -- directly and negatively affected by energy developments must have their concerns fully identified and sensitively addressed.” (Intro, Section 3, IL 89-4)

Amongst the initiatives that the Board “expects” & “encourages” industry to do are the following (Section 4):

(a) “Industry is encouraged to begin its public involvement and communication activities well before submission of any application to ERCB/AE, and also to establish effective two-way lines of communication with directly affected people prior to application submission.”

(b) “Industry should explore every avenue of problem solving available to it before bringing the unresolved concerns of citizens to an ERCB public hearing.”

(c) “Those sections of an application dealing with the community, social costs and benefits, the environment, human and animal health, and emergency preparedness should be framed at an easily understood language level.”

(d) “When preparing for an ERCB public hearing, the company should place considerable emphasis on communicating information directly related to the concerns of people and how those concerns will be addressed.”

The Board suggested, in effect, that the particular nature of a communications and public involvement program should reflect the nature of both the proposed facility and its possible risks / impacts, as well as the site-specific circumstances of the potentially impacted region.

There is no doubt that when it comes to public involvement, while the public has important rights, it also has certain obligations. The Guidelines (Section 3) emphasise this when they point out that there are a number of things that directly affected residents “could do”, whenever they have concerns about specific projects,

117 EMF Risk Perception and Communication in order to make the application process “more productive and less confrontational”. These include: attending public meetings and requesting small-group or one-on-one meetings with government and industry, bringing their concerns forward to the company involved or to the ERCB as early as possible, and attempting to resolve as many issues as possible prior to public hearing.

This type of regulatory Guidelines, whose influence in ERCB proceedings has been reinforced by subsequent Board decisions, have proven invaluable in proactively promoting public involvement programs which are so important in creating the type of calm, trusting and mutually respectful atmosphere that is conducive to an informed and constructive discussion and a co-operative resolution of risk issues. In this way, the Board’s Public Involvement Guidelines are also important not only because they help protect the rights of local residents and facilitate a proper review of scientific evidence but also because they are essential to advancing the broad public interest in Alberta as it relates to public confidence in both the energy industry and the regulatory system. Such public confidence is another pivotal pre-condition to the management of risk controversies.

CONCLUSION

The effective assessment and mitigation of potential environmental and health hazards can ensure that project-related risks are “low” and that the public will, in fact, be safe. However, such technical strategies alone cannot guarantee that concerned publics will also feel safe --- and thereby judge even such a low risk also to be “acceptable”. This second component of risk management, so crucial to earning public support and winning regulatory approval, requires the effective use of risk communications strategies, including meaningful public involvement programs, in the actual conduct and subsequent discussion of risk assessment, risk mitigation and other project planning activities. Risk communications is a key aspect of an integrated approach to the management of environmental risks as well as an effective strategy for avoiding and defusing EMF or other risk controversies.

118 EMF Risk Perception and Communication

Evaluating Response Options

Judy Larkin REGESTER LARKIN 16, Doughty Street GB-London WC1M 2PL, UK Tel: +44 171 831 3839, Fax: +44 171 831 3632

Introduction

By every objective measure - from infant mortality to life expectancy - westerners are healthier today and, arguably, exposed to fewer and lesser hazards than ever before. However, living longer and having few immediate, material concerns now seems to mean that we have more time to spend contemplating all sorts of long-term theoretical hazards. A catalogue of surveys demonstrates that we worry so much about the future that we believe new technologies and innovative products and services should not be permitted until it is known for certain that they won’t endanger our health or the environment. Just think what it would be like if coffee was discovered today. With 19 known carcinogens, it would provide worse test results than many banned synthetic chemicals!

So, while we’ve become increasingly informed, educated and opinionated – with rapidly rising expectations regarding consumer choice, access, efficiency and value, we’ve become much more anxious about the complexity and remorseless pace of technological and scientific change that both drives and serves these demands. Just looking in our UK newspapers over the last few weeks, we’re advised that Prince Charles doesn’t think soya is safe to eat, the discovery of dual- sex polar bears in the Arctic has renewed fears about synthetic chemicals, the safety of HRT has been challenged – again, the debate over genetically modified organisms continues, the findings of a report following the death of 21 people in Scotland in 1996 from the western world’s worst outbreak of E-Coli blamed a failure by environmental health experts to follow procedure……..and to cap it all we’re faced with the outlook depicted below!

EMF is a health issue which has the potential to create an unprecedented level of anxiety and controversy.

119 EMF Risk Perception and Communication

There are a number of key drivers which have shaping public perception of the EMF issue in the UK. An understanding of these is essential to the planning and evaluation of mechanisms for responding to the emotional dynamics associated with EMF, because the potential for public outrage looks set to escalate.

Although my presentation is based on experience of the issue in the UK, and there are important differences in the way in which the issue has developed country by country, the drivers identified have international relevance and the principles required to address public concerns are broadly transferable.

Other risk issues

Firstly, the background environment. A succession of risk issues in recent years has produced a cumulative effect on the public consciousness, typified by widespread cynicism of information provided by the industry and government, and with it have come demands for greater accountability, disclosure and public participation. Against a backdrop of more assertive consumer activism in our country, public scepticism and uncertainty grows every time a risk is perceived to have been mismanaged to the detriment of public health – from adverse health effects associated with asbestos, tobacco, infant formula, silicon breast implants and genetically modified organisms through to doubts over the safety of treatments such

120 EMF Risk Perception and Communication as hormone replacement therapy and the third generation contraceptive pill, and events such as the E-Coli outbreak in Scotland, the recent disastrous disposal of Brent Spar, and of course BSE.

Cumulative Effect of Past Events on Public Perception

Tobacco BSE

Brent Spar Asbestos

EMF E-coli Pill

GMOs HRT Silicon Breast Implants

The belief, at the height of the BSE crisis in the UK, that over a period of many years the public had been consistently lied to over the inadequacy of food processing standards and the availability of worrying research findings leading to a complete breakdown in institutional trust has fuelled widespread resistance to government and the industry reassurances couched as scientific or technical fact.

A recent primetime television news programme in the UK, compared the alleged health effects of mobile phones with both tobacco and BSE.

Issues such as these have produced a society predisposed to expecting adverse outcomes, no matter whether the risk is real or perceived, and one which is innately suspicious of organisations which seek to persuade otherwise. In the post BSE era the EMF issue therefore, could not have come at a worse time.

The UK press, one of the more competitive and intrusive media communities internationally, has been quick to identify a multitude of alarmist human interests angles potentially linked to EMF. And, not untypically, as you can

121 EMF Risk Perception and Communication see, alarming content about the possible risks dominates the headlines fuelling the outrage factor.

Media coverage

The UK has seen a significant increase in the amount of negative media coverage just this year. This has focused in the main on cellular phone handsets and the siting of masts, and has been characterised by sensationalist headlines, such as, “Will your phone kill you?” and “My mobile gave me cancer”.

Interestingly, the alleged health effects of handsets have received more coverage than those associated with masts, although there has been an increase in regional reporting of local citizens action groups protesting against individual mast sitings. This is significantly different to experience in other countries. In the Republic of Ireland the situation is more typical of experience internationally, with mast sitings much more prominent than handsets.

For our media, the possibility that mobile phone users, who have long been typecast as affluent, yuppie show-offs running their lives and everyone else’s on the phone in public places, should suddenly discover that the phones they are using could affect their health is richly ironic! Add to this the fact that the phones themselves have become symbols of the inexorable push of technological advancement and you have a cocktail of fear, uncertainty and disaffection.

122 EMF Risk Perception and Communication

The media is more than capable of exaggerating risk: alarmist messages make better stories than reassuring messages. Equally, traditional journalist criteria such as timeliness and human interest will carry more weight than the seriousness of a risk in technical terms. Coverage tends to focus on non-technical issues such as blame, fear and rage rather than quantification of the risk itself.

An analysis of the press coverage of the issue to date indicates a significant increase both in volume and in the amount of negative commentary.

Press Coverage on Hand Sets and Base Stations in the UK and Ireland to Date

7000000070 Million

6000000060 Million POS 5000000050 Million NEUT 4000000040 Million NEG 3000000030 Million 2000000020 Million

1000000010 Million 0 Audience Exposure

JAN FEB MAR APR MAY JUN JUL

A look at the broadcast coverage tells a similar and more potent story (video).

This growth in coverage is not just about an overall increase in airtime: this video indicates how scientists and activists have each received considerably more airtime than industry.

Both the press and broadcast coverage of this issue indicate that it is one which is constantly monitored by newsdesks. Articles and broadcast reports routinely include interviews with anti-industry activists who have been extremely successful in raising their profile, positioning themselves as credible experts and ensuring their voice is heard. The fact that a British tabloid newspaper recently printed a full page article by one such campaigner is indicative of the role of activists in escalating awareness of EMF health effects, and they represent another key driver of the issue.

123 EMF Risk Perception and Communication

Activists

A number of special interest groups have evolved around the EMF issue in the UK and these have become more sophisticated and better organised in their ability to communicate both with each other and with the media. In the UK the campaign against handsets has been spearheaded by high profile individual campaigners, while campaigns against masts, which are on the increase, have been organised by localised, but increasingly well organised, citizens groups. The video gave a taste of the range of people eager to communicate about this issue – from eccentric EMF boffins who conduct experiments in their homes, to well organised groups which publish newsletters and communicate with counterparts in the US and Australia, to anxious parents groups demanding a role in the planning process because no one can reassure them with any certainty that their children are and will be safe. All are highly motivated and many are extremely adept at attracting the attention of the media.

Proportion of Comment in Press Coverage from Industry vs. Scientists vs . Activists

23%

40%

37%

Scientists Activists Industry

Although not as sophisticated as similar groups in North America, press content analysis shows these UK groups are a force to be reckoned with. In the first eight months of this year, campaigners secured 37 per cent of total column inches, compared with just 23 per cent for the industry. As you can see, scientists have the

124 EMF Risk Perception and Communication loudest voice, with 40 per cent, and their role in shaping perception is therefore critical, but I will come on to that later.

In summary, public perception of the risks associated with EMF has been shaped, and will continue to be shaped by, the media, vocal anti-EMF campaigners and analogies with other risk issues. This is true of the issue both within the UK and internationally.

However, there are other, less tangible influences on public opinion which must be taken into account if the development of the EMF issue is to be fully understood and therefore managed to minimise public anxiety and maximise the effectiveness of communication initiatives, whether they come from the industry, scientists or government. These influences, or “emotional dynamics”, go some way to explaining why people do not react to risk issues in predictable ways, and also suggest why public consultation strategies can so easily fail.

Emotional dynamics

We know that when people make decisions about risk they do not simply make a clinical, machine-like response to factual information. Although we need, and are increasingly demanding, access to scientific information we are also strongly influenced by emotion and instinct, feelings of fear and familiarity, and a sense of having control.

This process is very difficult to quantify; risk means different things to different people, and some people will be prepared to take risks which others will refuse to consider. Attitudes to risk are often forged by a complex interplay of social and cultural factors, reinforced by friends, colleagues, family members and others. These attitudes are deeply embedded and hard to change, and they shape the way in which new risks are understood, interpreted and acted upon. Although this process is difficult to quantify, however, this model attempts to classify the dynamics which influence risk decisions.

We know that these dynamics are important in shaping behaviour: for example, any risk which carries a high “dread” factor, such as the possibility of cancer or a brain tumour, will always elicit a highly charged response. The fear people have of brain damage and madness goes a long way to explaining the panic over BSE. People will happily skydive in their leisure time or drive without a seatbelt during short journeys, while worrying about eating a tomato which has been

125 EMF Risk Perception and Communication genetically modified, and they will happily live along the San Andreas fault, accepting a risk they would never accept if it was linked to industrial activity.

The Emotional Dynamics of Risk Decisions

Awareness of risk Scientific Choice/control consensus RISK ACCEPTANCE Equity Nature

Detectability Dread

When you apply these emotional dynamics to the issue of EMF and mobile phones, the potential for public outrage – whatever the know consequences – becomes apparent. The prospect of a brain tumour of course carries a very high dread factor, as does the fact that EMF is not a tangible substance which can be detected in advance, or which can be measured in an easy or understandable way. In the case of handsets, the risk-taker enjoys a benefit, but this is not true of people living close to masts. Although there has not been outright disagreement amongst scientists on this issue, there have been a number of studies which are not easily comparable, and there has been dispute over the methodologies and validity of studies which lay people are unqualified to understand but which nevertheless have been highly publicised. There is a sense that we have been placed under a risk which has existed for well over a decade. Although mobile phone users choose to use their phones, people living close to masts in the UK have little say in the planning process and are now escalating their anti-planning campaigns. Finally, although EMF occurs naturally, radiofrequency generated by the mobile phone industry constitutes a man-made risk and is therefore more likely to be treated with fear and distrust, so these dynamics make the issue particularly hard to communicate.

126 EMF Risk Perception and Communication

Mobile Phones - Alleged Health Risks Handsets

people don’t know they may be at risk mobile lack of phone user scientific awareness of risk makes lifestyle consensus scientific choice/ choice /research consensus control RISK ACCEPTANCE equity nature risk is risk taker man-made enjoys benefit detectability dread

no warning of high dread factor brain tumour of brain tumour

And of course, lack of scientific certainty compounds the degree of anxiety. Scientists are placed in a unique position whereby they are trusted by the public and regarded as independent and credible sources of information, and as such they have a significant role in shaping the perception of this risk issue.

Science

In the absence of guarantees of zero risk people behave in a strongly risk- averse way and are easily influenced by agencies which they do not associate with the source of the risk: the media, special interest groups, scientific experts and regulators. Information which is associated with the organisation posing the risk will be treated with profound cynicism. For example, surveys have shown that people are much more likely to believe scientists working for pressure groups than they are scientists working for industry. Among all these groups independent scientists command most credibility – they have the voice which people are least likely to consider partisan. They therefore have a crucial role in shaping people’s perception of risk and the onus is on them to balance the information flow from the media and campaigners with their own communication.

127 EMF Risk Perception and Communication

However, the polarisation of attitudes to risk which exists among experts and members of the public poses a problem in communicating individual risks which goes beyond the science of the issue concerned. Scientists can sometimes find it hard to understand why society is unable to come to terms with the uncertainties inherent in all research while members of the public may feel that experts traditionally take too narrow a view, dealing with statistics rather than taking on board the significance of possible personal consequences for individual people.

In addition to this, scientists may be frustrated by the fact that, although governments, regulators and commercial organisations may share their concerns about the way in which members of the public seize upon data, public anxiety may reach a level whereby they are unable to postpone policy or business decisions until more research is complete.

Public Confidence in Scientists Working for..

97%

77% Environmental Groups Environmental 64% Government Industry

Source: Mori 1995

These issues do not make the role of the scientist seeking to assess the level of risk an easy one. However, their views are in constant demand.

In the UK at present, the demand for information from scientists has been met by a steady flow of individual studies: in recent months we have seen the Mild,

128 EMF Risk Perception and Communication

Braune and Bastide studies, as well as research published by the Defence Evaluation Research Agency in the UK and considerable publicity around the work conducted by Dr Alan Preece at Bristol University. However, there has been a tendency on the part of the media to misinterpret such studies – and this has not been addressed properly by the scientific community. The Mild study on handset use and headaches, published in May, was rapidly picked up by the media following a press conference in Stockholm and was fuelled by simultaneous publicity over a first test case in the UK relating to a mobile phone user with a brain tumour. The Braune study, which found a link between changes in blood pressure and mobile phone usage, received widespread coverage yet only one newspaper in the UK reported that it was based on just ten people. It is difficult for members of the public to evaluate the significance of these studies or to make an informed decision about their importance when they lack an understanding of the way in which they were conducted and are reliant on media interpretation. In the light of this the responsibility for allowing them to make informed decisions lies with the scientists who conduct studies and then publish their results. An added burden I know but it is essential that the utmost care is given to ensuring the right information is available, and inaccuracies are corrected when they appear in the media.

That is why the WHO programme, which has provided a framework by which to catalogue studies and assess their significance, is so important. It has been invaluable in providing information on when scientific assessments can be expected, and this has been important in deflecting allegations of secrecy. The credibility of the programme is reflected in the fact that it has been increasingly referenced in media coverage over recent months.

The final key driver which has a key role in shaping public perception is the industry itself. Given our backdrop of profound media scepticism and public cynicism this is a challenging process. A survey by a leading research in the UK found that only 15 per cent of people thought multinational companies were trustworthy. More specifically, a recent survey amongst consumers conducted in a mobile phone trade magazine indicated that as many as one in five mobile phone users – that’s over one and a half million people - are cutting down on their calls because of health fears – so the commercial impact cannot be ignored and the industry must take urgent steps to be seen to act in a co-ordinated and responsible fashion.

The media analysis and video I showed earlier highlighted that the mobile phone industry has a much smaller voice that either the scientific community or

129 EMF Risk Perception and Communication activist groups. It is imperative that the industry adopts a more proactive stance with the media, by delivering consistent information and correcting inaccuracies. To gain credibility the industry must be seen to know more about this issue than anyone else. Failure to do so will have damaging commercial consequences for a high growth international industry and for a host of technical applications that offer real benefits to consumers.

Looking at a lifecycle model, this issue has now moved well into the current phase, awareness has reached a level whereby there is considerable pressure on both science and industry to quantify the alleged risk and rebut analogies with the tobacco industry and BSE. The potential for litigation against the industry is now high, and demands for tighter regulation of the industry will become more frequent, with a real likelihood of changes in standards guidelines looming.

Issue Life Cycle

Opportunity to Influence Difficult to Influence

Formal Constraints

Period of Increasing Awareness

Media Coverage Pressure

Potential Emerging Current Crisis Dormant Development

If communication continues to be led by the media and activists, with inaccurate reporting of scientific research going on unrebutted, industry may

130 EMF Risk Perception and Communication ultimately be forced to change the way they operate regardless of whether or not a risk is proven.

The fusion of all the elements I have outlined today which are driving this issue forward in the public consciousness may prove so potent that it could become the most significant risk issue the UK has seen to date.

The mobile phone industry is dynamic and innovative, offering significant benefits to consumers and we should not lose sight of this. There is a real opportunity and need for industry, the scientific community and government agencies to move forward together, supporting the WHO and other independent research frameworks, providing clear, accurate and consistent information to consumers in order that they can be reassured by responsible action and make informed decisions.

131

EMF Risk Perception and Communication

The Internet as a Public Information Resource, with a case study on a Canadian controversy about radio-frequency fields

William Leiss, F.R.S.C. Eco-Research Chair in Environmental Policy, School of Policy Studies, Queen’s University Kingston, Ontario, Canada K7L 3N6 Tel: (613)545-6832, Fax: (613)545-6630 E-mail: [email protected]

with Greg Paoli, M.A.Sc. President, Decisionalysis Risk Consultants, Inc., 76 Park Avenue, Suite 302, Ottawa, Ontario, Canada K2P 1B2 Tel: (613)565-5722, Fax: (613)565-5941 E-mail: [email protected]

Abstract:

Installations for four of the new digital telecommunications networks for personal communications services (PCS) began in Canada in 1997. These networks utilise the 2 GHz part of the spectrum, whereas the older cellular telephone technologies use frequencies in the 800-900 MHz range. One significant feature of these new networks is that they require numerous roof-top antenna installations. In Canada one federal department, Industry Canada, issues licenses for transmitter and antenna installations, while another, Health Canada, is responsible for the health and safety (risk) regulations, which are published as “Safety Code 6” (issued 1990). Safety Code 6 specifies exposure limits for workers and citizens and covers frequencies ranging from 10 kHz to 300 GHz.

In the summer of 1997 some citizens in Vancouver, British Columbia became aware of the new installations when their community was informed that a local school had been asked to allow the placing of a roof-top antenna on its building. Shortly thereafter, the citizens discovered that another antenna had already been installed inside the steeple of a nearby church; the church hosts a day-care centre in its facilities. Concerns about health risks were raised, public meetings (including meetings with representatives of the two federal departments) have been held, and the controversy has been simmering ever since.

An important part of subsequent events was that the citizens turned to the Internet for information and self-education in health risk issues of concern to them. We examined these resources, and this paper describes the results of a detailed search of the Internet on hypothesised human health effects of radio-frequency

133 EMF Risk Perception and Communication electromagnetic fields. A conceptual map was developed which captures linkages between concepts according to their proximity and interconnectedness in the Internet milieu. A key issue is the abundance of summaries of scientific information provided in lay terms by non-institutional sources. These summaries tend to include very frank discussions of key concerns in scientific and regulatory inference and frequent reference to the rationality of the burden of proof applied in science where public health protection is concerned.

This paper provides an account of the public controversy, down to the present, from the standpoint of the need for good risk management principles as well as good risk communication practices.

INTRODUCTION.

In late April 1998 Canada’s national newspaper, The Globe and Mail, published a front-page story entitled “How the Net killed the MAI.” MAI is the Multilateral Agreement on Investment, a proposed international agreement on investment rules which has been championed by OECD countries for the last few years. A consortium of advocacy groups around the world, opposed to this initiative (partly because it was perceived to be mainly in the interest of multinational corporations), used Internet resources to co-ordinate their campaign, creating attractive web sites, maintaining constant communication, and distributing key information instantaneously among their memberships, including leaked official documents. One noteworthy aspect of the campaign is the fact that allies in Third World countries can participate at very low cost. In the context of the announcement that the OECD had suspended its efforts to reach agreement among official government representatives, one diplomat commented: “This is the first successful Internet campaign by non-governmental organizations. It’s been very effective.”7 In our opinion, it will not be the last.

One of the major arenas for advocacy-group intervention in global public policy issues is in the broad area of health and environmental risks, where the “official” players at the table are these same national governments and multinational corporations. For some time now organizations such as Greenpeace and the World Wildlife Fund have been active at a global level in matters such as persistent toxic chemicals, forest management, and wildlife preservation, utilising competent scientists on their own staffs. Among other things, they have developed effective Internet web sites as a part of their campaigns.

7 Madelaine Drohan, article in The Globe and Mail, 29 April 1998, p. A1.

134 EMF Risk Perception and Communication

One can expect that the more successful environmental advocacy groups will become increasingly skilled in their use of Internet resources to advance their causes. But as well, and increasingly, individual members of the public who do not necessarily have scientific expertise, but who have concerns about risk issues, have begun using Internet resources to gather information, establish contact with like- minded people everywhere on the globe, obtain guidance on how to ask questions of experts, and prepare themselves to become skilled intervenors in risk controversies. The information-search, documentary retrieval, and networking facilities of the Internet have huge advantages over earlier resources available to the general public, and these advantages will grow steadily in future years. From now on, all major controversies over risk management decision-making will be played out on a public stage with an international cast and audience.

In addition to the aspects already mentioned, such as ease of networking and relatively low cost (including the value of time), the World Wide Web has other important characteristics considered as an information resource. One is the ease with which linkages across related concepts and areas of interest can be tracked, especially types of linkages which very likely would not have occurred spontaneously to individuals, nor would they have turned up in the cross-indices of the older library card catalogues. Another is ease of access to a wide variety of very different information sources, many of which are interactive -- that is, an individual’s queries will result in personal responses and offers of further assistance. These are two of many similar advantages responsible for the success of the Internet. These and other features can have very important impacts on the formation of connections between ideas, on the apparent plausibility of cause-and-effect relations, on the almost imperceptible blending of scientific and anecdotal evidence, and on the basic framing of risk issues. All of these features represent new challenges, discussed later on in this paper, for those who are responsible for the social management of risk issues.

In this paper we concentrate on the implications, for risk communication and risk management, of the use of the Internet as an information resource for individuals and groups in controversies over health and environmental issues. Our case study is the recent and ongoing controversy in Canada over health risks associated with the radio-frequency (RF) fields that are utilized in wireless communications technologies.

Conceptual Map of Internet Coverage for RF/EMF Issues.

135 EMF Risk Perception and Communication

It is very difficult to describe the nature of the information on the World Wide Web in a linear format on paper. Our diagram8 is an attempt to demonstrate the enormous ability of the Web to capture complex phenomena in ways that normal documents cannot. The diagram is based on an intensive survey of the Web carried out in a purely random, experiential mode. The connections indicated by arrows create a map of the thematic linkages which can be “discovered” by anyone who is browsing the Web. This is by no means an exhaustive representation of all of the possible links that may be followed. In other words, ours is only one of many such maps that could be drawn, and the network of connections as a whole is in fact far more complex than this diagram suggests.

The figure also reflects the position of health effects within the larger thematic map. Health effects are sandwiched between technology (at the bottom of the diagram) and the societal variables which drive the public determination of the acceptability of risks and the required interpretation of uncertainty. As an example of the connectivity of issues, the following example describes the path through various concepts as might be followed by browsing the Web.

One web page contains an excerpt from a 1987 editorial in the New England Journal of Medicine, suggesting more cautionary approaches to regulatory toxicology;9 this thought leads to a discussion of the “orthodox” nature of science and the apparent inability or “refusal” of the scientific and regulatory communities to accept the possibility of adverse health effects beyond those known as thermal effects. This leads to a discussion of the inadequacy of current exposure limits and the fact that Soviet researchers have lower limits based on their belief in the possibility of non-thermal effects. Often, commentators will question why the burden of proof is on the government or the public.

Interestingly enough, most sites do not insist with certainty that there are adverse health effects from RF exposure beyond those described as thermal effects. However, there is considerable discussion of “prudent avoidance,” which is

8 The diagram will be found at the end of the paper. For a sample of what may be found on such sites, go to: http://www.cruzio.com/~rbedard/waveguide/library.html 9 “Science is a hard taskmaster, and in the light of mounting evidence that suggestions of toxicity are for the most part ultimately confirmed by painstaking scientific inquiry, perhaps it is time to reexamine whether scientific standards of proof of causality – and waiting for the bodies to fall – ought not to give way to more preventative health policies that are satisfied by more realistic conventions and that lead to action sooner.” (19 April 1987)

136 EMF Risk Perception and Communication described as a common-sense precaution in the face of uncertainty.10 Here it should be noted that the doctrine of prudent avoidance is often supported on the web-sites of companies which provide exposure mitigation devices (such as antenna shields) or consultancies in how to measure the home environment for those who are “electrosensitive.” Another site featuring prudent avoidance is maintained by an independent researcher in the U.K. who has launched a lawsuit to require hazard labelling on cell phones; this same researcher offers on the site an elaborate theory that homosexuality, AIDS, and Sudden Infant Death Syndrome (SIDS) are a result of damage to the brain in the area that controls the immune system.

“Electrosensitivity” is a much more mature issue in Europe, and in particular in Sweden, than it has been so far in North America. The “electrosensitive” contributors on the Internet Europe usually refer to “electrosmog” and then make the link to a general concept of pervasive environmental decay resulting from the actions of insufficiently controlled industries.11 The general environmental decay theory is linked to other feared exposures to widely dispersed contaminants and is further linked to any health impact that is, or is perceived to be, increasing in incidence in the population. Correlation between radio-frequency exposure and increases in brain cancer, breast cancer and asthma are frequently cited, with the asthma hypothesis supported by some scientific evidence of interference with the activity of antihistamines.

Usually, the discussion returns to what is perceived as denial or cover-up; parallels to the tobacco case are almost always provided, and some sites have exhaustive discussions of allegations about a long history of denial about RF effects – by the U. S. military, for example.12 This leads in some cases to a discussion of the military/industrial complex and the sheer economic power and momentum of the mobile communications industry. The economic power is in turn seen as explaining how these impacts could be covered up; sometimes the cover-up is linked to the use

10 See below, pp. 22ff., for further discussion of prudent avoidance applied to RF issues. 11 In European countries the term “electrosmog” is a catchy shorthand phrase to denote the sum total of electrical and magnetic fields (EMF) disseminated through the operation of all communications-related technologies – cellular phones, short-wave radio, radio and television broadcasting, radar and other microwave installations. “Electrosmog” is not yet a prominent part of the lexicon of public controversy in Canada. See P. M. Wiedemann & H. Schütz, “The Electromagnetic Fields Risk Issue,” European Review of Applied Psychology, vol. 45 (1995), pp. 35-39. 12 The history is chronicled in Paul Brodeur, The Zapping of America: Microwaves, Their Deadly Risk, and The Cover-up (New York: W.W. Norton & Company, 1979).

137 EMF Risk Perception and Communication of RF in brainwashing experiments by the CIA. Finally, RF and power-line (ELF or extremely low frequency fields) issues are closely paralleled on the Web, in particular with regard to the provision of resources (by grassroots activists linking communities with each other) to fight the siting of various types of installations. Any health effect in ELF is a candidate in RF and vice versa; for example, adverse health effects such as Alzheimer’s and ALS in the ELF domain are also featured very strongly in the RF domain, and both are linked to findings on memory loss, CNS effects, and some so-called “subjective” effects reported by Soviet researchers in past decades.

“Electrosmog” in Canada.

Apart from a minor level of interest in the older controversy about cellular telephony and brain cancer, there had been no sustained controversy in Canada over RF fields until four networks (Cantel - AT&T, Clearnet Communications, Microcell Telecommunications, and Mobility Canada) began to install the newer digital personal communications services (PCS) networks, first in the larger urban areas, in 1997. While the older analog cellular telephones operate at 800-900MHz in the high frequency band, the PCS systems occupy the ultra high frequency band (around 2GHz); this higher frequency allows the latter to operate at very low power. The output powers of the base-station antennas are on the order of 100 watts each, and the handsets themselves have a maximum output power of around 1 watt. However, the low-power mode requires a clear line-of-sight between the antennas in the network, which means that a comparatively large number of antenna installations is necessary (750 antenna sites in Metro Toronto, 500 in Montréal, and 500 in Greater Vancouver).13

Telecommunications regulation is a federal jurisdiction in Canada, and licenses for the transmitter and antenna sites needed for cellular telephony networks are assigned by Industry Canada. Antenna licenses for cellular systems are issued in groups of 80 to cover a defined geographical area, but particular sites are not specified. Licensees are required to comply with a set of safety guidelines issued by another federal agency, the Radiation Protection Bureau of Health Canada, and also to undertake consultations with “land-use authorities” (municipalities); obviously, consent of a property owner also must be obtained before an installation can be

13 Clearnet PCS, Inc., “PCS: What is it?” pamphlet, December 1997; Ericsson Radio Systems AB, “Health and Safety in Mobile Telephony,” pamphlet, 1997.

138 EMF Risk Perception and Communication erected.14 Except where violation of a valid municipal zoning ordinance (such as height restriction) is at issue, the Industry Canada license cannot be legally challenged providing that consent of the property owner has been obtained. Not surprisingly, the perceived lack of community and local government control over such installations, especially in “sensitive” locations (e.g., on schools and churches), is itself a factor in public controversy.

In early June 1997 parents in the Fraserview area of Vancouver learned that Cantel had made an arrangement to install a PCS antenna on the roof of the local elementary school building. An outcry erupted at a school information meeting and Cantel quickly cancelled the deal; however, as a result of the publicity the community discovered that Microcell was in the process of installing a 1.2-metre antenna inside the cross atop the Fraserview Assembly Church located across the street from the school building.15 Microcell refused to cancel this project, and some community members began to mobilize public opinion against it, preparing a news release which was picked up in newspapers and granting TV and radio interviews. Over the next months protest leaders prepared pamphlets for distribution in the community, wrote letters to papers, held meetings with local government officials, and filed an appeal against the installation with the Vancouver Board of Variance, which deals with zoning matters.

A Health Canada official had been present to discuss health risk concerns at the initial information meeting, but within a week newspapers were citing local school board officials as saying that New Zealand had banned such antennas on school buildings, and the states of Oregon and Washington were considering similar actions. The protest leaders began calling the scientific staff at Health Canada headquarters in Ottawa with requests for more details about the scientific evaluation of risk; a prominent researcher in this field at the University of Washington was also

14 Radiation Protection Bureau, Health Canada, “Limits of Exposure to Radiofrequency Fields at frequencies from 10kHz – 300GHz [Safety Code 6],” 1991; revised draft, April 1997. Industry Canada, Spectrum Management, “Environmental Process, Radiofrequency Fields and Land-Use Consultation,” 1995. At least some provincial governments, such as British Columbia, maintain staff with technical expertise in EMF and RF, but this level of government does not have an official role in siting decisions. 15 Church steeples are a preferred site for wireless telecommunications equipment locations, and in the United States some consultants are making hefty fees advising church congregations on how to bargain effectively with the industry representatives. Jon G. Auerbach, “Steeple chase a serious business for U. S. churches,” The Globe and Mail, 1 January 1998, p. B5. Cf. Kevin Marron, “Antenna on church is a test of faith,” The Globe and Mail, 2 September 1997.

139 EMF Risk Perception and Communication quoted in early media coverage, warning that “there has not been enough research for scientists to know if there are cumulative long-term effects on children from these antennae.”16 As a result of direct contacts with Health Canada staff, university-based researchers, and their first Internet searches, the protest leaders began to inform themselves about the scientific risk assessments for radio-frequency fields.17

Within six weeks the issue was tabled in a formal setting (the Board of Variance meeting) with a structured mode of confrontation. Microcell filed an information package about scientific and regulatory matters with the Board in advance, and brought to the Board’s public meeting not just its own personnel but also a university professor well-known in this field, who was acting as a consultant to the company.18 The Microcell representatives made presentations and were asked questions by the Board members. The community protest leaders who had filed the appeal were there too, of course, led by Mr. Milt Bowling and Ms. Angela Sousi; they had filed a written brief of impressive proportions with the Board and made oral presentations, the nature of which tell a great deal about the forms of information traffic in the Internet age.

The reasons for appeal in the citizens’ brief and oral presentations, which had been researched and written within about six weeks, ranged over regulatory and scientific matters and were based on personal discussions with some academic researchers in the field as well as on articles in a leading industry publication, Microwave News.19 Bowling and Sousi opened the “credibility” gap by contrasting what they were being told – about the scientific studies (including animal studies) on which the current risk assessments are based – unanimously by health experts from

16 Professor Henry Lai, quoted in the Vancouver Sun, 10 July 1997. 17 At about the same time (24 June 1997) the city council in the neighbouring community of Burnaby heard a submission against a re-zoning application to permit construction of an antenna in the parking lot of a shopping centre. The submission was made by a Mr. J. A. Whiffen, who described himself as an equipment designer for the nuclear diagnostics industry. Whatever the merits of the particular case being advanced, the submission presents what is on its face a highly sophisticated argument by someone thoroughly conversant with the scientific and technical terminology. 18 The consultant, Professor Maria Stuchly of the University of Victoria, has a special significance in this controversy, because at an earlier stage in her career she was the federal government employee who was chiefly responsible for preparing Safety Code 6. 19 Traditional print-based publications are still important in the electronic age: Microwave News was a key source for the citizen intervenors in the early stages of their investigations into the health risk issues.

140 EMF Risk Perception and Communication their own local, provincial and federal governments, by the company, and by the company’s own academic expert, with references to different views apparently held by some Harvard University researchers and by Michael Repacholi, the leader of WHO’s EMF risk program. Safety Code 6 was said by them to be out of date, being based on the state of research in the 1980s and not having been revised to reflect any of the numerous research results published in the preceding ten years.20 Of course the Vancouver Board of Variance was in no position to take up matters such as the possibility that recent studies had shown excess lymphomas in laboratory mice exposed to RF fields; in any case without giving reasons the Board concluded the meeting by revoking the construction permit for the antenna installation in the church.21

As mentioned earlier, the federal jurisdiction is paramount in these matters, and Microcell decided not to remove the installation (the church did not require it to do so), which in effect threw the issue back into Industry Canada’s lap. For the citizens who had led the fight the effect of Microcell’s perceived intransigence was to prolong the issue, and they responded by redoubling their efforts to prevail. As the issue persisted through the Fall of 1997, the role of the WHO program and its director increased in prominence. In October Angela Sousi remarked on a Vancouver radio program:22

[Our] concerns are based on the fact that we feel the scientists in the world can’t unanimously agree on the safety of the long term non-thermal bioeffects of non-ionising radiation and … when we had a meeting at our school earlier this year, we were told that it was impossible to have non-thermal bioeffects, meaning that if it doesn’t heat you, it can’t hurt you. There was a study done recently that indicated

20 Safety Code 6 is officially a guideline, not a regulatory instrument; federal officials maintain that this is actually an advantage, since guidelines can be changed far more easily than regulations. However, it does have the character of a “static” document, since no new material appears in the booklet itself since its original issuance in 1991. With the single exception of a 1990 publication which does not deal with health issues, there are no publications dated later than 1988 in the list of references in Safety Code 6. 21 One week earlier (16 July 1997) another municipal official, the Medical Health Officer at the Vancouver/Richmond Health Board, had issued a one-page sheet supporting Health Canada’s position: “… [T]he general scientific consensus holds that the power from cellular base stations is far too low in the community to result in adverse impacts.” There was also a very pertinent comment on prudent avoidance, to which I shall return later; but this document does not appear to have exerted any influence on the course of the debate. 22Interview with Rick Cluff, “Early Edition” program, CBU-AM (Vancouver), 24 October 1997.

141 EMF Risk Perception and Communication that is a possibility that we could have non-thermal bioeffects. This was conducted by a doctor, Michael Repacholi, on behalf of Telstra [the Australian national telecommunications company],… [who] did some experiments with other individuals that suggested that there were some effects. So here we have one group of scientists saying it’s possible and another saying it’s impossible.

Expert disagreement – and the impacts it can have on the public perception of risk -- is not unknown in risk controversies, including all types of EMF risk.23 The impact it may have under specific circumstances can vary greatly. In this case, it seems to us, the sequence of events and the nature of the actual players made a difference to the outcome.

In the present story one key event was a press conference in Geneva held on 19 December 1997, at the conclusion of scientific meetings sponsored by WHO and attended by experts from seventeen countries, at which Michael Repacholi called for more research on EMF and RF risks.24 The lead paragraphs in the wire story ran as follows:

Hoping to sort out “mixed evidence” on the issue, the World Health Organization called Friday for more research into whether mobile phones, power lines and radar might cause diseases that include cancer and Alzheimer’s. Dr. Michael Repacholi, manager of the WHO’s Electromagnetic Fields Project, told a news conference that perceived risks from new technologies have become a serious public health issue. He expressed confidence that existing international standards are adequate for high-level exposure, but said study is needed on the effect of low- level exposure over longer periods.

In the story as a whole there were references to “suggestions,” “possible connections,” “possible links,” and “mixed evidence” between EMF at various frequencies and the following list of diseases: brain cancer, head and neck cancers, childhood leukaemia, lymphoma, breast cancer, central nervous system disorders, memory loss, neuro-degenerative diseases, and Alzheimer’s. On mobile phones in particular, there was a direct quote from Repacholi, to wit:

23 W. Leiss & C. Chociolko, Risk and Responsibility (McGill-Queen’s University Press, 1994), chs. 4-5. 24 Citations in this paragraph are from the nearly-identical stories published in many western Canada newspapers on 20 December, all based on a Reuters wire story from Geneva.

142 EMF Risk Perception and Communication

“Mobile phones have only been around for less than 10 years now and the incubation period for cancer is at least 10, maybe 15 years. So we need to set up the studies so that if there is an impact, they [sic] can be found in a reasonable time.”

The public might be excused for thinking, upon reading this statement, that ordinary citizens differed from laboratory rats chiefly in that the latter were used for short-term experiments and the former for the longer-term ones. More to the point, perhaps, the citizen intervenors in Vancouver might have been excused for thinking that the substantive health-risk basis for their bitter and prolonged fight against PCS antenna siting had been validated fully by an eminently credible international agency.

Risk Communication and Risk Management challenges.

In 1998 Microcell “voluntarily” removed its antenna installation from the Fraserview Assembly Church in Vancouver. And in August 1998 the Government of Canada asked the Royal Society of Canada to establish an independent expert panel on the health risks associated with radio-frequency fields; the panel will issue a public report in March 1999.25 Within hours of the press release announcing the

25 Press release, Royal Society of Canada, Ottawa, 4 August 1998. The Terms of Reference for the panel report ask the panel to answer the following questions: A) General Questions: With regard to Health Canada's Safety Code 6, in particular the draft version revised as of July 1998: 1) What are the biological effects and/or potential adverse human health effects associated with exposure to radiofrequency fields emitted from wireless telecommunication devices such as wireless phones and base-station transmitters? 2) Do the provisions of Safety Code 6 protect both RF workers and the general population from the “thermal” effects associated with the exposure to radiofrequency fields? B) With regard to the issue of "non-thermal" effects of radiofrequency fields: 3) What "non-thermal" biological effects and/or potential adverse health effects have been reported in the literature? 4) Is there evidence that such effects, if any, could be greater for children or other population sub-groups? 5) What research is needed to better understand the potential health consequences for "non-thermal" effects? C) Implications of the foregoing for Safety Code 6: 6) What are the implications for Safety Code 6 of the panel’s scientific review of the currently available data on biological effects and the potential adverse health effects

143 EMF Risk Perception and Communication panel’s composition and mandate, the Society began receiving a continuous stream of return messages, including copies of queries sent by Internet activists to a variety of lists asking for an analysis of the panelists, with respect to their “positions” on issues and on sources of their research funding. Two of the panelists immediately were said to be compromised by close affiliation with the wireless telecommunications or electrical industries.26

There is every reason to think that we are just at the beginning of the period when Internet resources will be used by individuals and groups to enhance their ability to be skilled intervenors in social controversies about health and environmental risks. The radio-frequency fields controversy in Canada allows us an opportunity to do an initial assessment of the role of Internet resources in these matters, within the broader context of society’s own need for appropriate risk management processes. For it is one thing to observe that Internet-based information resources are vital new aids in the empowerment of citizens, and thus in the functioning of legitimate democratic decision making processes – which is the case, generally speaking, in my opinion. But it is equally true that sensible risk management is a complex high-stakes and long-term game in which we are obliged to reflect carefully on the assessments and perceptions of risks, as they evolve in protracted risk controversies, and to ask, for example, whether individuals and groups actually are serving their own interests in the outcomes. Thus the real heart of the matter is not the uses of Internet resources themselves, but rather it is the relation between these uses and the ability of individuals to promote their own “best interests” in the positions they take on how our society should manage risks.

The underlying structure of risk controversies is best understood as a state of legitimate and necessary tension between the expert assessment of risk, on the one hand, and the public perception of risk, on the other. Because the gap between the two is an enduring rather than a transitory phenomenon, we require a means of

of exposure to radiofrequency fields? In particular, should the phenomenon of “non-thermal” effects be considered in Safety Code 6? 26 The author is Chair of the Society’s Committee on Expert Panels which screens and appoints panelists and oversees the procedures by which panels operate. The Society has published a “Manual of Procedural Guidelines” (October 1996) for panels which is based on procedures developed by the U. S. National Academy of Sciences and its affiliated institutions; these include provisions for a detailed review of “bias” and conflict of interest, and an effort to achieve “balance” in panel composition, in the course of the panel selection process. Of course, whether the Society has achieved these objectives in any particular case is a legitimate subject for debate.

144 EMF Risk Perception and Communication facilitating exchanges between the two domains, and this is in fact the function of good risk communication practices. More often than not, however, we find a systematic failure to employ such practices, leading to the creation of a risk information vacuum which makes sensible and publicly-credible risk management decision making virtually impossible to achieve.27

The radio-frequency fields controversy in Canada is a perfect case study in the existence of a risk information vacuum and its consequences for risk management. This can be seen clearly once the elements presented earlier are arrayed systematically. In summary, there is clear evidence that governments had authorized private industry (in return for financial considerations) to introduce a new technology across Canada, including numerous installations at sensitive locations within communities, without first having in place a clear and credible explanation of the associated risk factors. This is in our opinion undeniably a dereliction of duty for governments, which have the primary responsibility for managing risks.28 From a practical standpoint it is also asking for trouble. One hopes that both industry and governments have learned something from this venture.

We have organized the issue map into four sectors: risk communication, risk assessment, related “social” issues, and risk management.

Risk Communication Issues.

1. Insufficient explanation of technical and engineering factors: The full electromagnetic spectrum, and its radically different characteristics at various frequencies, is a complicated business. In addition, the trade-off between the frequency and the output power of transmitters (where higher frequencies make it possible to use lower output and thus – arguably – a net reduction in total risk, considering multiple risk factors) is directly relevant to perceived risk. Another example: The shape of the beam from a base station antenna has some specific characteristics

27 For a full discussion of this perspective see chapter 1 in Douglas Powell & William Leiss, Mad Cows and Mother’s Milk: The Perils of Poor Risk Communication (McGill-Queen’s University Press, 1997). 28 This is just one in an almost endless list of such cases, where governments think they have finished their work after their risk assessment is completed, having devoted little or no effort to risk communication, including confronting the matter of credibility – and then usually only after a controversy has erupted, when it is virtually impossible to get one’s message across.

145 EMF Risk Perception and Communication

directly relevant to exposure, and thus to risk;29 among other things, this is relevant to the application of the “prudent avoidance” maxim (see below). But none of this was explained in advance to the public in understandable terms. Once the controversy broke in a specific community in Vancouver, government and industry personnel scrambled to provide some explanations for these and other technical factors; but one cannot hope for much success in persuasive communications under such conditions. (A good time to initiate the dialogue with affected communities would have been when the installation siting was in the planning stage. But this would violate the time-honoured maxim, “Let sleeping dogs lie.”)

2. Explanation of risk factors, (a): federal government – basic document. Safety Code 6 is an engineering-type document, containing nothing but the barest mention of the health risk factors. Most significantly, it does not even name the established risk factor (thermal effect), much less any of the hypothesised, so-called non-thermal effects; thus it cannot serve the purpose of risk communication. Nor does a small Health Canada brochure with the odd title, “Safety of Exposure to Radiofrequency Fields: Frequently asked questions,” mention any specific risk factor. The brochure states: “We hope it provides clear information about a complex and often misunderstood topic.” Since the text mentions the possibility of adverse effects, but does not even say what they are, the intended clarity is a vain hope.30

3. Explanation of risk factors, (b): federal government – current science. Everyone who accesses the Internet on RF issues quickly learns about “non-thermal effects,” but (so far as we know) no document issued by

29 “The signal from a cellular or PCS base station antenna is essentially directed toward the horizon in a relatively narrow beam in the vertical plane. For example, the radiation pattern for an omni-directional antenna might be compared to a thin doughnut or pancake centered around the antenna while the pattern for a sector antenna is fan-shaped, like a wedge cut from a pie…. Consequently, normal ground-level exposure is much less than exposure very close to the actual antenna.” U. S. Federal Communications Commission (FCC), Office of Engineering and Technology, “Information on human exposure to radiofrequency fields from cellular and PCS radio transmitters,” January 1998, p. 2. 30 The FCC has useful and up-to-date documents on its web site that are written in a reasonably accessible style (www.fcc.gov/oet/rfsafety), especially OET Bulletin No. 56, “Questions and answers about biological effects potential hazards of radiofrequency radiation.”

146 EMF Risk Perception and Communication

the Canadian government has ever used these words. The published scientific literature includes mention of the following types of such effects (note that any such effect may or may not be substantiated or that, even if it is, an effect is not necessarily an adverse effect in human health terms): brain cancer; other cancers; calcium ion efflux; ocular damage; electrosensitivity; stress; birth defects; headaches; asthma; immunosuppression and immunostimulation; alterations in drug metabolism; memory loss; behavioural changes; learning deficits; Parkinson’s disease; Alzheimer’s disease; ALS.

The federal government, elected by Canadians, has decided to authorise industry to install technologies in return for handsome payments; information in wide circulation (not all of which is suspect by any means) associates these technologies with certain risk factors. Why is it thought to be appropriate for public authorities in Canada to remain silent on these matters? Canadian citizens could, if sufficiently motivated, turn to the web site maintained by the U. S. Federal Communications Commission and at least read a brief discussion on non-thermal effects, and learn that at least two of them (the “calcium efflux” and “microwave hearing” effects) are regarded as well-substantiated. They would also find the following general statement: “It is possible that ‘non-thermal’ mechanisms exist that could cause harmful biological effects in animals and humans exposed to RF radiation. However, whether this is the case remains to be proven.”31 But why should they have to go to a foreign government for such a discussion, however minimal?

1. Explanation of risk factors, (c): industry. (i) Microcell official quoted in a newspaper: the antenna poses “no risk.” (ii) Same official: “There isn’t a safety concern.” (iii) Cantel official: “There’s absolutely no medical or scientific research that indicates any cause for health concerns.”32 The following comments are pertinent: (i) Of course it poses a risk – although the risk may be negligible, vanishingly small, acceptable, or whatever. (ii) Of course there is a safety concern – although that concern could be reduced with more complete or reliable information. (iii) Of course there is such research – although it may not turn out to be substantiated. It is surprising that such officials still do

31 Ibid. (OET Bulletin No. 56), p. 5. 32 (i) Vancouver Sun, 10 July 1997; (ii) )The Province (Vancouver), 2 July 1997; (iii) The Vancouver Courier, 8 June 1997.

147 EMF Risk Perception and Communication

not seem to realise that such statements are both inaccurate and inflammatory.

2. Timeliness. The lack of advance discussion with respect to the installation of PCS network equipment is a serious risk communication failure, as already indicated. But the next generation of wireless communications technologies is already in the works; are these mistakes likely to be repeated? The terms of reference proposed to the Royal Society of Canada for its review of health risk factors associated with RF fields contain the following statement: “New technologies such as mobile data, wireless local area network (WLAN) in the 5 GHz range, specialized mobile radio (SMR)/enhanced specialized mobile radio (ESMR), wireless local loop and low earth orbit (LEO) mobile satellite service are forthcoming.”33 Indeed, sooner than many may suspect. A press release issued recently states: “Canada’s largest advanced, wireless, broadband telecommunications network will be rolled out in all regions of Canada, beginning in Toronto in the first quarter of 1999…, [providing] connectivity for data, Internet, voice and video traffic over the air that can be delivered worldwide. The WIC Connexus advanced network represents the genesis of the 28GHz wireless access market.”34 These so-called “local multipoint communications system” (LMCS) wireless networks eventually may put the transmitter/receiver equipment in everyone’s back yard. This may very well be just fine; but it would be wise to start talking about it with the public without delay.

Risk Assessment Issues.

1. Expert uncertainty and disagreement. Such matters ought always to be freely disclosed, assessed, and interpreted for the public by the risk managers. There is of course nothing surprising in the back-and-forth among researchers, especially where a relatively new area of research (such as non-thermal effects of RF fields) is concerned. This would be a matter of idle academic curiousity were it not for the fact that governments are collecting monies for allowing others to generate these fields while the research effort proceeds, which gives rise to the corresponding responsibility for regular reporting and discussion.

33 Supra note 19. 34 WIC Connexus press release, Toronto, 19 August 1998.

148 EMF Risk Perception and Communication

2. Quality of the basic document. Some of the shortcomings of Safety Code 6 as a risk assessment document have been indicated earlier (supra note 14). A revised version has been under development for some time but is not yet available, partly because Health Canada’s Health Protection Branch does not have nearly enough resources to cover its responsibilities. But the document format also has many deficiencies, especially in that it treats its subject-matter primarily as a matter of engineering rather than of basic science, and in that it does not explicitly refer to risk assessment (RA) methodologies or the incorporation of RA into a “formal” risk management process.

“Social” Issues.

1. Equity: This is always fundamental where siting is concerned, and since siting choices are left up to industry, it is industry who should deal with them. Communities are entitled to hear reasons in favour of particular siting choices and an answer to objections to those choices – without having to insist on having them.

2. Prior notice. This too should be provided without citizens having to ask for it, but it certainly did not happen with the Fraserview Assembly Church site. Both citizens and municipalities (which are first in line for complaints from their residents) deserve complete disclosure for the entire installation plan in their areas. In November 1997 the Municipality of West Vancouver’s general services committee approved recommendations for mandatory notification to Council of PCS sites (even when approval is not required), a public notification and information process by companies prior to issuance of permits, and an overall plan for all sites to be requested by a company, so that alternatives could be discussed.35

3. Role of governments. Both Industry Canada, which has legal authority to approve telecommunications installations by private industry all across Canada, as well as the various provincial and municipal authorities which are passive partners in these ventures, should be far more sensitive than they are now to the potential for community outrage. This is especially so since the actual authorization for licenses

35 Corporation of the District of West Vancouver, Council Report: Personal Communication Service (PCS) Antenna Cell Sites Approval Process,” 28 November 1997.

149 EMF Risk Perception and Communication

takes place in Ottawa, far from the localities where the impacts will be felt. The advent of the next generations of wireless technologies offers an opportunity for governments and industry to collaborate in designing an enhanced community liaison process.

Risk Management Issues.

1. Allocating responsibilities for enhanced risk communication: Where several agencies of government, in collaboration with industrial sectors made up of many different players, introduce new technologies and new risks, it is unclear who should be assigned responsibility for risk communication. The usual result is that no one picks up the ball. This should be sorted out before the crisis strikes. All these players may be perceived to have some conflict of interest with respect both to the risk assessment and scientific research overview as well as the risk communication activities, but this difficulty can be addressed by using independent third parties for such functions.

2. Uncertainty. The most problematic aspect of risk management from an ethical and equity standpoint – although it has never been adequately recognised as such – lies in justifying who in society should bear the costs of uncertainties in risk assessments.36 In the earlier history of innovative risk-taking in industrial society those who could not avoid unequally distributed and often excessive exposures (workers) paid the price for the fact that new technologies were implemented for long periods while massive uncertainties about the associated hazards persisted. Others not so exposed reaped the greatest share of the benefits, but the rising standard of living for everyone also distributed widely a considerable share of the benefits.

Things have changed considerably on this score. Generally speaking, societal risk management has reduced the excess involuntary risk for sub- populations (e.g., those living in proximity to hazardous waste facilities, RF antennas, or nuclear power stations) to a mere fraction of its former level. Another way of putting this is to say that the benefit – risk ratios in involuntary exposures have widened enormously to the advantage of the former. Those risk levels are not and never can be zero; but they are not necessarily an unreasonable burden on those

36 Risk and Responsibility, ch. 9.

150 EMF Risk Perception and Communication who now bear them, taking into consideration all of the known risk factors in the lives of people in contemporary society.

But the excess risk (however small) in every involuntary exposure itself, and the ever-changing elements of persistent uncertainty in the risk assessments, and the rationale for inequitably distributed exposure, must be articulated and defended to the citizenry by the risk promoters. (Risk promoters are industry and governments which introduce new technologies.) Such a defense may vary widely in terms of its rationale, but it can be constructed, and in any democratic society worthy of the name, the risk promoters have the duty to do so.37 That duty rarely has been discharged at all, much less discharged well, in our own society, and it is past time to change this pattern.

1. Risk reduction. In all cases where unequally-distributed involuntary exposure is combined with a mixture of private benefit and public good, risk reduction to some defensible point (say, “as low as reasonably achievable”) is a sound and just risk management option. In EMF and now in RF issues the phrase “prudent avoidance” has wide currency in this context. It implies that the technology’s social benefit is evident and that exposures should be minimized so far as especially sensitive populations such as children are concerned. This application of the maxim of prudent avoidance was articulated clearly by Milt Bowling: “We want sites located where children aren’t going to be exposed for long periods of time every day.”38

Ironically there is the possibility that the maxim could be applied best by choosing the site which was first the subject of protest, namely, the roof of the local school building. This has to do with the “shape” of the beam in a RF directional antenna (see note 23), which would result in virtually zero exposure in the building itself as well as its surrounding grounds. This point was made by Microcell’s expert consultant at the Board of Variance meeting, and also in the circular prepared by the local Medical Officer of Health:

37 This is not the same as demanding consent from everyone who might be exposed at any level, because no new technology ever could be introduced under such a requirement, and on the whole, and collectively, citizens derive significant net benefits from them. 38 Cited in Kevin Marron’s article in The Globe and Mail (2 September 1997) and in other news stories.

151 EMF Risk Perception and Communication

With respect to the particular installation and the specific request to the Board of Variance to institute a moratorium on cell phone antennae sites in the vicinity of schools and day cares, the benefits of a moratorium are at best questionable and at worst non-existent. Given the typical radiation patterns from cellular antennae, there is normally a “radiation shadow” directly beneath the antenna structure with very low levels (well below 1µw/cm2) of radiofrequency radiation. Most of the antenna power is directed outward horizontally (within a 10° cone @ 100 ft) usually commencing at a height of several stories above ground level. The practice of “prudent avoidance” in this instance does not result in any increased level of protection as might be the case in requiring buffer zones next to high voltage transmission lines (where both magnetic and electric fields are present as opposed to RF fields).39

This is an example of an important point where citizen intervenors themselves need to be very clear as to their own interests and objectives, first of all, and secondly, as to the risk management options that will serve those objectives well. The prudent avoidance maxim itself is not necessarily irrelevant to the RF fields issues; what is at stake here is whether or not it can be applied to achieve some sensible risk reduction objective and if so, how.

In the end the responsibility to propose such applications – more precisely, to initiate a dialogue with citizens in which the possibility of applying the prudent avoidance maxim sensibly to particular circumstances is raised -- does not lie with worried parents, but rather with the risk promoters. In other words, when the industry initiates -- as it should -- a community dialogue early in its own planning process for new installations, a part of its own contribution to that dialogue is to demonstrate that it has explicitly and seriously reviewed its preferred siting options from the standpoint of the prudent avoidance maxim.

Conclusion on Risk Management: A Note on SwissRe.

One of the most interesting documents yet to appear on RF issues is the booklet “Electrosmog – a phantom risk,” published in 1996 by Swiss Reinsurance Company (Zurich). The document is also posted on the company’s web site and thus can be downloaded anywhere, and this is how it was obtained by some of the citizen intervenors in Vancouver, who had seen a reference to it in Microwave News. It deals in a highly sophisticated way with the business -- and specifically the

39 See note 16.

152 EMF Risk Perception and Communication insurance -- risk posed by the existence of social controversy over health effects associated with EMF and RF fields, which in SwissRe’s view gives rise to

… an extremely dangerous risk of change composed of two parts: the classical development risk, that is, the possibility that new research findings will demonstrate electromagnetic fields to be more dangerous that has hitherto been assumed; and the sociopolitical risk of change, in other words, the possibility that changing social values could result in scientific findings being evaluated differently that they have been so far…. We consider the risk of change to be so dangerous because it is evident that a wide range of groups have great political and financial interest in electrosmog being considered hazardous by society. If these interest groups prevail, current and future EMF liability suits could be decided in favour of the plaintiffs, thereby confronting the insurance industry with claims on a scale which could threaten its very existence…. In this sense, this publication is a warning.40

This warning comes from an industry that is still paying out asbestos-related claims for exposures that occurred up to half a century ago. It carries the implication that liability insurance coverage for the telecommunications industry could be affected by the further development of these issues in “sociopolitical” terms.

Perhaps the most curious aspect related to this publication is that at least some of the citizen intervenors in Vancouver who read SwissRe’s booklet found succour in it: They thought that it validated their expressions of concern, even though this very carefully written document gives no support to the idea that there are now unacceptable levels of risk associated with RF fields, or that they are likely to be found unacceptable in the future. What can explain this reaction? True, the document has beautiful graphics, including a superb one on the EM spectrum; it has a most intelligent and readable scientific summary of EMF; it takes the matter of persistent uncertainty seriously; it concedes the point that “it is theoretically possible for even the weakest signal to induce biological responses and in this way affect organic processes” (p. 17); it rejects the false comfort of zero risk, stating clearly that “EMF health risks cannot be eliminated entirely … [but] can at best be reduced, insofar as they are known and measurable” (p. 20); and, above all, it has a sophisticated discussion about the relation between scientific risk assessment and the

40 “Electrosmog – a phantom risk,” pp. 4-5 (italics in original). Recently another global player in the forest industry, Canada’s MacMillan Bloedel, conceded that changing social values about clearcutting and old growth forests had affected significantly its business risk .

153 EMF Risk Perception and Communication social evaluation of risk. But none of this probably explains the reaction of some of the readers in Vancouver. We suspect that their reaction reflects, perhaps subconsciously, a sense of appreciation for the fact that a major industry player had taken the trouble to write about these things at all.

Strengths and Weaknesses of the Internet as an Information Resource.

Earlier we noted that individual members of the public who have concerns about risk issues have begun using Internet resources to gather information, establish contact with like-minded people everywhere on the globe, obtain guidance on how to ask questions of experts, and prepare themselves to become skilled intervenors in risk controversies. Those who wish to do so can find on the Web complete copies of many peer-reviewed scientific publications and other documentary material from excellent sources. The information-search, documentary retrieval, and networking facilities of the Internet have huge advantages over earlier resources available to the general public, advantages that will grow steadily in future years.

There are some corresponding disadvantages as well. Many sites are maintained by activists who are committed to a particular perspective on issues and who also have (judging by the contents) reasonably good scientific training; individuals visiting these sites who are non-experts in these matters can end up just with a wider array of opinion on what the scientific issues are, without any way of evaluating the relative merits of what they find. Second, much (but by no means all) Internet activism has the tenor of “guerrilla warfare” and conspiracy, a crusade against the large institutional players in government and industry, which influences the presentation of material. Third, straightforward scientific reports are mixed liberally with anecdotal evidence; casual visitors to their sites need to exercise some caution in sorting through what they find.41

These and other weaknesses are serious matters, but on the whole they do not cancel out the offsetting advantages. In any case the Internet as a public information resource is here to stay. Citizens concerned about health and environmental risk issues will derive greater benefit from Internet resources over time as more players set up shop there – including those who have a mission to

41 The author’s e-mail address was given on the press release from the Royal Society of Canada; within days there was a regular return flow of communications about RF issues, containing among other things detailed cases about alleged adverse health effects for both humans and farm animals in Australia, Canada, Germany, and the former Yugoslavia.

154 EMF Risk Perception and Communication deliver balanced, disinterested, up-to-date, and credible accounts of ongoing risk controversies.

Prepared by Decisionalysis Risk Consultants

Scientific Lack of Preparation Expectation of Editorial, Environmental Aesthetics Statement by Orthodoxy in Evolution for Conspiracy NEJM, 04/87 Awareness NEMFAC this Exposure & Cover-up Chairperson

“Science is Blind Soviet Research to Non-Thermal Lack of Local Ignored “Electronic General Effects” Smog” Environmental Decay Empowerment

Industry Military Inadequacy Bias in Burden Profit Motive Applications of Proof Unrestrained & Secrecy of Current Technology Exposure Limits Proliferation Chronic Environmental History of Lawsuit re: Denial of Labelling of “Why Not Exposures Impacts Phones (U.K.) Prudent Avoidance?”

“20th Century Increased Incidence High-Power Nuclear Companies Promoting Diseases” “Tobacco” Exposure Mitigation in Health Impacts Radar Systems Radiation Devices & Services

Environmental Increase in Endocrine Asthma Immune Illnesses Cancer Effects, CIA Mind Suppression Fertility Control & Brainwashing “Sick-Building Electro- “Subjective Microbial Brain Breast Syndrome” Sensitivity Illnesses” Leukemia Infections Cancer Cancer Memory Loss

Interference Pacemakers Melatonin Alzheimer’s ALS Various CNS Effects Interference Hearing Aids Neurochemical Mechanisms (BBB, Calcium) Occupational Handsets Residential Antennas School Exposure Exposure Children

Home VDTs Power AM/FM & Radar Radon X-Ray Nuclear Nuclear Appliances Lines TV Reactors Weapons

ELF RF IR

155 EMF Risk Perception and Communication

156 EMF Risk Perception and Communication

Public Debate on Electric and Magnetic Fields and Health in Quebec

André Beauchamp ENVIRO-SAGE Inc., 25 rue Jarry ouest, Montréal, Québec, H2P 1S6 Tel: (514)387-2541, Fax: (514)387-0206

INTRODUCTION

While the first public debate on electric and magnetic fields (EMF) and human health took place in 1975 during hearings held by the National Energy Board (NEB) of Canada in conjunction with a proposal to export energy to the United States (Cardinal, 1996), it is, above all, through the public hearings of the Bureau d’audiences publiques sur l’environnement (BAPE) that this issue has been repeatedly examined in Québec

This article focuses on the examination of the problem during five public hearings before the BAPE involving Hydro-Québec, the promoter, in respect of five different projects. I will look at the increasingly important role that the health sector is playing in the debate and how, through the interplay between Hydro-Québec, the BAPE and the health sector, future developments can be anticipated. In light of these observations and my experience, I will formulate a number of suggestions concerning risk management and EMF.

OVERVIEW OF THE ISSUE

Certain projects in Québec, including the construction of 351 kV and higher power transmission lines are subject, by regulation, to an environmental impact assessment. Under the process, the promoter must issue notice of the project, the government department concerned must issue a directive concerning the environmental impact study, and the promoter must produce the study under the supervision of the Minister of the Environment and Wildlife. Once the impact study has been completed, it is made public. Any individual or institution wishing to request the holding of public hearings may do so within 45 days. Unless the request is deemed to be frivolous, the minister must grant the hearing and mandate the BAPE to conduct it.

The BAPE has four months to carry out its mandate to conduct an investigation and organize public hearings. While the procedure is quasi-judicial in nature, public hearings in Québec are political and popular in nature. The hearings are held in two stages. The first stage is devoted to questions and the dissemination of information and the second, to the submission of briefs. At the conclusion of its

157 EMF Risk Perception and Communication deliberations, the BAPE presents its conclusions and analyses in a report. The body does not have decision-making power but plays an advisory role, i.e. it engages in non- binding arbitration. Cabinet makes the final decision, on a recommendation from the Minister of the Environment and Wildlife. In order to exert some influence in decision- making and maintain its credibility with the public, the BAPE tends to draw attention to its conclusions.

Hydro-Québec is one of the biggest promoters in Québec. It is a government corporation with very close ties to the government. Established in 1941, Hydro-Québec earned US$5 billion in 1995 and has become a symbol of the Québec identity and a key lever in the government's economic strategy. A love-hate relationship has developed with respect to the utility: it is admired but vehemently criticized as well (Paquet, 1988). The deregulation of the energy market and Hydro-Québec's determination to export energy to the United States also complicate the situation. For the past 20 years, no enterprise has done as much as Hydro-Québec in the realm of impact assessments and with respect to a knowledge of the territory. At the same time, the utility's projects invariably arouse strong resistance and lively criticism.

FIVE PUBLIC HEARINGS

Between 1983 and 1996, Hydro-Québec appeared successfully on five occasions before the BAPE. The Great Whale project was also subject to a detailed environmental impact assessment, under a federal and provincial commission encompassing five jurisdictions. However, the utility withdrew the project before public hearings were held.

1983

From June 17 to October 17, 1983, the BAPE held public hearings on a proposal to build a 735-kV power transmission line running above all through farmlands, a converter station, and a "450 kV line for an interconnection with the New England states. During the hearings, Quebecers strongly contested Hydro-Québec's position, on the strength of recent studies by Wertheimer and Leeper, and demanded that the utility take a much greater interest in health issues. The farming community also expressed considerable anxiety about the effect of power transmission lines on animal health. The BAPE commission expressed the opinion that current knowledge did not warrant halting the project or imposing a moratorium on it. However, it did request that Hydro-Québec contribute more extensively to research and suggested that epidemiological studies be carried out.

158 EMF Risk Perception and Communication

When the BAPE authorized the project on March 28, 1984, the Québec government demanded that Hydro-Québec conduct an epidemiological study on human health in collaboration with the ministère des Affaires sociales (MAS), which later became the ministère de la Santé et des Services sociaux (MSSS), and the ministère de l'Environment (MENVIQ), now the ministère de l'Environnement et de la Faune (MEF), and a study on the health of livestock, in collaboration with the ministère de l'Agriculture, des Pêcheries et de l'Alimentation (MAPAQ) and the ministère de l'Environnement.

1987

From December 8, 1986 to April 8, 1987, the BAPE once again held hearings on a "450 kV line. During the hearings, Hydro-Québec explained that the epidemiological study requested was undergoing a feasibility analysis. Moreover, in its directive concerning the impact study, the MENVIQ did not deem health to be a significant factor from the standpoint of the location of the line. Consequently, it simply requested information documents on the question. The most striking intervention during the hearings was the brief submitted by the Département de santé communautaire of the Centre hospitalier de l'Université Laval (DSC-CHUL), presented by three health experts, i.e. Pierre Lajoie and Patrick Levallois, physicians specializing in community health, and Éric Dewailly, a physician and toxicologist. In response to this highly critical brief, the Commission took the MENVIQ to task for attaching little importance to health, criticized Hydro-Québec for its lack of concern with health, and recommended that the MAS establish a committee to assess the value of Hydro- Québec's studies on health and EMF. On June 10, 1987, the Québec government set up the committee requested to monitor Hydro-Québec studies on electric and magnetic fields. It also encouraged Hydro-Québec to take into account human exposure to EMF when reviewing design and operating criteria.

1992

In order to pursue its deliberations, the interdepartmental committee called upon the services of the DSC-CHUL. In 1991, three members of the DSC, Patrick Levallois, Pierre Lajoie and Denis Gauvin, published a report on the state of knowledge of health and EMF (Levallois, Lajoie and Gauvin, 1991). Two experts from Hydro- Québec also collaborated. The publication of the report heightened awareness in the Québec medical and public health sectors of EMF and human health.

159 EMF Risk Perception and Communication

From October 30, 1991 to February 29, 1992, a new BAPE commission examined another Hydro-Québec 735-kV power transmission line project. The commission did not hire a health expert and the DSC did not submit a brief. However, the MSSS delegated Patrick Levallois and Denis Gauvin to represent it at the hearings. In light of documents submitted by Hydro-Québec and Patrick Levallois's testimony, the commission did not object to the project on health grounds but recommended that the interdepartmental committee assume responsibility for overseeing the studies and that a feasibility study be conducted in respect of epidemiological research on the level of exposure of the population to EMF and the latters' effect on health. Furthermore, the commission referred to the Prudent Avoidance mentioned in the report by Levallois, Lajoie and Gauvin (1991) and recommended that Hydro-Québec bolster its information strategy.

1993

From March 29 to July 29, 1993, a different BAPE commission examined another 735-kV transmission line project, which was to cross an agricultural and tourist zone with high potential because of the scenery. The MSSS delegated to the hearings three experts, i.e. Dr Marc Dionne, Dr Patrick Levallois and Dr Denis Gauvin. The DSC submitted to the commission the report by Levallois, Lajoie and Gauvin (1991). The commission devoted an important chapter of its report (68 pages) to the question of health and EMF and published a review of the literature prepared by the DSC-CHUL, research projects, the Québec plan of action on EMF, and the list of epidemiological studies then under way (document submitted by the utility). The commission examined the question at length. It first summarized expert opinions that had appeared between 1990 and 1993, the position of the interdepartmental committee and Hydro-Québec management, described EMF and took stock of studies devoted to health, i.e. experimental and epidemiological studies. The commission then reviewed the mitigation measures suggested by various studies. It expressed the opinion that posing the problem in terms of health and technical and economic advantages was insufficient. "The decision on whether or not to develop and apply mitigation measures in respect of electromagnetic fields should be taken in light of general public protection guidelines and satisfy public expectations concerning security and the quality of life" (BAPE: 263) (our translation). The commission pointed four possible solutions: the adoption of a standard, prudent avoidance, the status quo, or a moratorium. The commission opted for prudent avoidance and requested the elaboration of a mitigation strategy, the reassessment of transmission and distribution technologies (the designing of power lines that produce less electric and magnetic fields, the designing and manufacturing of electrical appliances that produce low EMF, and the training of electricians). The

160 EMF Risk Perception and Communication commission recommended the adoption of mitigation measures when new lines are located and the supervision of research and technological development to reduce exposure to EMF. In order to implement what it called "prudent management," the commission requested the reorganization of the interdepartmental committee and its application to the Hydro-Québec development plan.

From a critical standpoint, the commission's position seems paradoxical. The commission deemed the health effects to be simply plausible. It mentioned the concept of prudent avoidance, submitted to it by Patrick Levallois, although it did not analyse it in light of the writings of Granger Morgan and did not discuss it at length. However, in practice, in my opinion it went well beyond prudent avoidance and proposed, under the term "prudent management," a fundamental change in planning and management, with a view to mitigating exposure to EMF. The commission temporarily overlooked scientific uncertainty and requested the implementation of a prevention policy. This stance was a political choice aimed at protecting the public and satisfying public expectations concerning security and the quality of life. It seems evident that the members of the commission expected the numerous studies then under way to demonstrate beyond a shadow of a doubt the impact of EMF on health.

On May 29, 1994, the follow-up committee, in the wake of an examination of the CHUL report and the analysis of the findings of research conducted by Dr Gilles Thériault, EDF-OH-HQ, acknowledged that "certain doubts exist about the cause relationship between high-voltage lines and some cancers" but that scientific data and results do not warrant "recommendations to the effect that changes be made in the design and operating practices used for electrical facilities." However, "it seems fitting that H-Q, the utility and its partners examine ways of implementing prudent avoidance with respect to new electrical facilities" (Cardinal, 1996: 4).

Hydro-Québec then initiated research on prudent avoidance. It concluded that the concept of prudent avoidance is inconsistent from a conceptual standpoint, and inadequate. It is inopportune for medical, ethical, legal, engineering and communications reasons. H-Q defined its strategy, which it called "prudent management" and regarded as more attractive than prudent avoidance, although not the prudent management mentioned in BAPE report 68 (H-Q, 1996). Confusion surrounding the terms is astonishing: there are two different concepts of prudent avoidance and two different concepts of prudent management.

1996

161 EMF Risk Perception and Communication

From April 23 to [?] 19, 1996, another BAPE commission examined the Duvernay-Anjou 315-kV power transmission line project, which was to cross a near- urban agricultural zone, Rivière-des-Prairies, and a completely urban zone in the Montréal Urban Community. Strong opposition to the project was voiced. The participants opposed the choice of the route and wanted the line to be located somewhere else. They opposed the construction of an overhead power line and demanded an underground line for the urban section of the route and the portion crossing Rivière-des-Prairies. Opponents were deeply concerned about health risks.

During the hearings, the commission devoted an entire session exclusively to the question of health. Three specialists, Dr Louis Drouin, a physician with the Direction régionale de la santé publique de Montréal-Centre (DRSP), Dr Gilles Thériault, a researcher at McGill University and the director of extensive research on the health of workers in the electricity sector, and Paul Héroux, also from McGill University, represented the MSSS at the hearings. The MSSS emphasized uncertainty and the impossibility of concluding that a cause-effect relationship exists. During debate on the matter, Dr Michel Plante from Hydro-Québec expressed the opinion that the cause-effect relationship was simply plausible. Dr Thériault confirmed that this relationship is probable. A participant submitted at the hearing a videocassette of an interview with Dr Thériault on the French-language CBC network. When asked if he would buy a house near a high-voltage power line, Dr Thériault answered No, especially out of concern for his family's health. The question was obviously loaded since it confused expertise and personal attitude. If the expert answers No to this kind of question, he is confirming that there is a risk; if he answers Yes, he is undermining his own credibility. Opponents frequently cited Dr Thériault's testimony in their brief.

During the hearings, the commission received 28 briefs, of which 15 referred to health, all of them confirming that EMF affect health. Ten of the briefs requested, among other things, that power lines be buried. Some witnesses adopted a rhetorical stance when dealing with health issues, while others deemed such issues to be more important. Simply stated, the argument ran as follows: doubt exists; in the face of doubt, it is essential to abstain; therefore, ….

During the hearings, the RRSP submitted a brief signed by Dr Louis Drouin and Dr Gilles Thériault. In light of the review of the literature conducted by Levallois and Gauvin (1994) and an unpublished synthesis of few studies by Thériault and Chung Yi-Li (1996) on the risk of leukemia, the brief concluded that:

162 EMF Risk Perception and Communication

• the proposed power transmission line did not pose a health risk for people then living along the route;

• the users of a linear park planned along the line were likely to be exposed to fields of over 2 mG;

• because Thériault's studies tended to demonstrate excessive risk of leukemia among individuals exposed to more than 2 mG, it would be advisable to analyse the hypothesis of setting an exposure limit of 2 mG along new lines in respect of day care centres, playgrounds, schools and homes.

BAPE Commission 107 is closely monitoring the RRSP brief. It deems Hydro- Québec's rejection of the concept of prudent avoidance and the utility's proposal concerning prudent management to be equivalent to maintaining the status quo. Without analysing in detail Hydro-Québec's arguments, the commission believes that a prospective and preventive approach must be advocated. The commission rejects Hydro-Québec's concept of integrated development, which allows for the use of the right-of-way for recreational purposes, one of the main benefits for the public. Moreover, the commission has requested a moratorium on the concept and has requested that the proposed 30-m right-of-way be widened, in light of the 2 mG (0.2 µT) value suggested by the RRS. "We are facing considerable uncertainty concerning the procedure for action and the scope of the direct and indirect risks for human health of EMF. Everything therefore become conditional and perception and apprehension take the place of truth" (BAPE 107: 137) [our translation]. According to the commission, perception figures with respect to five factors: inadvertent risk, scientific uncertainty, potential risk for children, the invisible nature of EMF, and the spectre of cancer. The commission has asked the interdepartmental committee to examine the timeliness of establishing an exposure standard, given the wide margin between the Hydro-Québec, IRPA (100 µT) and RRS (0.2 µT) standards pertaining to magnetic fields. Furthermore, the commission recommends burying the line without discussing the impact of such a decision on magnetic fields.

The government has approved the construction of the Duvernay-Anjou overhead line, as requested by Hydro-Québec, without the broadening of the right-of- way and without restriction on the use of the right-of-way.

Since then, the January 1998 ice storm has revealed the fragility of Hydro- Québec's system under extreme conditions and appears to warrant extensive

163 EMF Risk Perception and Communication

Year Project Hydro-Québec Health BAPE Government 1983 735 kV (Plante) C Step up research Authorizes project "450 kV Simplified summary Conduct studies on health Requests feasibility study 1987 "450 kV (Cardinal) DSC-CHUL Denounces MENVIQ health Authorizes project Update brief (Lajoie, directive Levallois, Inter-departmental Dewailly) Calls for inter-departmental committee (MSSS) committee (MSSS) (Contract DSC- CHUL) 1991 C C Publication of CC Levallois, Lajoie, Gauvin report (1991) 1992 735 kV Updating of MSSS: Request to broaden mandate Authorizes project summary of Levallois, of inter-departmental studies Gauvin committee (MSSS) Refers matter to inter-departmental Request for epidemiological committee feasibility study

Request that H-Q bolster information 1993 735 kV (Goulet, Pineau) MSSS: Summarizes knowledge Authorizes project Updating of Dionne, summary of Levallois, Requests political intervention Refers follow-up to studies Gauvin MSSS committee Requests reorganization of committee 1996 315 kV (Goulet, Plante) DSC: Drouin, Criticizes H-Q's prudent Authorizes project as Document on Thériault, management request by H-Q prudent manage- Héroux ment Criticizes H-Q's integrated Refers follow-up to development MSSS committee

Recommends prospective and preventive approach

Recommends consideration of exposure standard

Recommends burying the line intervention by the utility in order to bolster the system. Consequently, the government has authorized two 735-kV and one 315-kV lines while suspending the environmental impact assessment procedure and adopting a streamlined procedure. The individuals concerned perceive these rapid decisions by the government as a strongarm tactic. Critics suggest that the stabilization of the system also masks an electricity exporting strategy. Debate over the effects of EMF on health has been vigorously revived, all the

164 EMF Risk Perception and Communication more so as the procedural fairness of the decision seems dubious or at least a big step backward in relation to previous cases. In the absence of fairness, the perception of health risks is changing. Once merely plausible, such risks are now at least regarded as probable, if not uncertain. It is hard to predict at present further developments in this respect.

REVIEW OF 15 YEARS OF DEBATE

An analysis of 15 years of debate on the question of electric and magnetic fields elicits, in my view, the following observations:

• The BAPE has been the focal point of public debate on EMF. From one hearing to the next, the media display almost no interest in the issue. During public hearings, debate resumes among participants in the hearings and the specialized media. Until now, debate during the hearings has not received prominent coverage in the media and has scarcely affected public opinion.

• Debate on EMF and health during BAPE public hearings has led, first and foremost, to the pursuit of research in this field. In light of BAPE reports, the government has put pressure on Hydro-Québec to fund new studies, constantly take stock of studies under way and bolster its international presence. It is reasonable to think that it is the possibility of a new examination of a proposed power transmission line before the BAPE that is compelling Hydro-Québec to pursue its studies, produce interim reports and formalize its strategy.

• In the realm of EMF and health, the BAPE's reports, while they have had little effect on public opinion, have forced the government to intervene constantly in this regard. The matter has progressed through a follow-up committee. There is a growing discrepancy between the BAPE's recommendations and the government's decisions.

• Over the past 15 years, it is the health sector (MSSS, DSC, RRSP, follow- up Committee) that has displayed genuine leadership in this field. It has exerted pressure to obtain resources. It has made accessible to the public scientific studies under way and disseminated the concept of prudent avoidance. It has heightened awareness in the public health sector of the possible effects of EMF on health. The BAPE's recommendations have

165 EMF Risk Perception and Communication

been inextricably linked to the suggestions of experts form the Québec health sector (DSC, MSSS).

• Health experts have played a number of different roles, including that of representing the health sector, i.e. as independent experts, before the BAPE, and as interveners, i.e. militants, through the submission of briefs during public hearings. This situation can be explained by the limited number of experts available in Québec and the disproportionate resources available to Hydro-Québec and experts on EMF in relation to those available to the health sector.

• In the course of 15 years of debate on EMF and health, almost always the same experts have appeared on behalf of Hydro-Québec and the health sector. Some 15 experts seem to have formed a club and to repeat their performances from one hearing to the next.

• While Granger Morgan, who conceived the notion of prudent avoidance, constantly refers to the calculation of the costs involved, this aspect is almost totally absent from debate on EMF in Québec. Granger Morgan deems prudent avoidance to be a concept to manage uncertainty over the very existence of a risk. Levallois, Lajoie, Gauvin (199) give the concept a broader meaning, approaching prevention. BAPE reports 68 and 107 emphasize this trend. While acknowledging uncertainty over the existence of risk, the BAPE's recommendations suggest that such risk must be regarded as established, either because it is probable or because the public perceives it to be so. BAPE reports 14, 22 and 47 clearly consider uncertainty to be too great to impose constraints. The reports emphasize research and development and knowledge. Reports 68 and 107, while they recognize uncertainty, insist on the urgency of strategic modifications and the consideration of EMF during the design phase of projects, as if their effect had been demonstrated. The recommendations largely exceed the concept of prudent avoidance. The last two BAPE reports (68 and 107) clearly attach a great deal of importance to studies under way, which should confirm the harmful effects of EMF on health. Has the weight of this foreshadowing been confirmed? Here, we are shifting from the analysis of completed studies to the outcome of studies under way or planned.

166 EMF Risk Perception and Communication

• It strikes me that there is an obvious difference between the private deliberations of the follow-up committee, including the issuing of opinions on health matters, and the prospective suggestions made publicly to the BAPE by the same health experts. The opinions on health matters maintain that "in the current context, we do not have at our disposal sufficiently conclusive scientific data and results to recommend changes in existing practices respecting the planning and management of electrical facilities" [our translation]. During their testimony, health experts have called for preventive management, the elaboration of a standard, or the prohibition of certain activities. This suggests that the health experts change their viewpoint depending on whether they are discussing the matter among themselves or expressing themselves in public. Scientificity prevails when experts discuss matters among themselves, and health protection predominates when they discuss matters in public.

• In the course of public discussions, we never have access to the actual studies, which are much too complex and difficult, but to reviews of studies and, occasionally, reviews of reviews. Since each review is also a reading and an interpretation, it is hard to sort out the matter.

• A militant who participates in a public hearing to protest the routing of a power transmission line or to reject the project itself tends to view risk as being established. Through rhetoric, uncertainty surrounding risk turns into certainty on risk. From one debate to the next, the perception of risk tends to become stronger. From the standpoint of social dynamics, debate over EMF and health constantly eludes the experts and becomes a social issue. A few vague indications are all that is needed for an issue to take on a social cast. There is an overriding impression that the government would like to close the door and refer the matter to experts, while the public formulates its own discourse on the limitations of expert knowledge.

• Research concerning perception and communication on risk tends to reveal that risk management cannot rely solely on the scientific measurement of risk, which is never entirely value free, without simultaneously taking into account perception and, consequently, the manner in which the populations concerned define risk. The procedure suggested by the Presidential- Congressional Commission (1997) is probably the most appropriate. Debate in Québec, and a steering committee that I chair, are moving in the direction. For the time being, the main field of application contemplated is

167 EMF Risk Perception and Communication

that of the contaminated soils policy, where it seems easier to weigh the benefits and drawbacks and control measures. Can this type of procedure be applied to the question of EMF when there is uncertainty over the very existence of risk? In my view, it can. In Québec, this would demand either a review of the environmental assessment procedure or the implementation of a credible, virtually permanent parallel procedure among the various partners. Some examples are a citizens' panel or a permanent committee, which would enable the citizens concerned to become familiar with EMF.

• In light of current uncertainty and the difficulty of drawing conclusions, will research simply be abandoned in the absence of convincing results with respect to a seemingly limited number of effects? Will research continue and if so, in what manner? Or, will a systematic prevention policy be adopted? Given the current situation in Québec, unless the international community and, above, the United States, exert heavy pressure, or unless evidence is obtained rapidly, it is unlikely that the government will compel Hydro-Québec to adopt a vigorous mitigation or prevention policy. The burying of power transmission lines in urban areas will probably serve as a mitigation measure, despite the expense incurred and the dubious effectiveness of the measure in terms of EMF. The exemplariness argument will probably prevent the government from ordering the widening of the right-of-way or the integration of health issues into the planning process. Such decisions would involve enormous cost for an uncertain benefit.

• With regard to perception, if the public fails to once again get involved, we are likely to enter a stalemate, partly because of emerging uncertainty about the threat of cancer and leukemia and because the routing of a power line can always be contested. It would be sufficient for a child living near a power distribution or transmission line to die of leukemia to crystallize public opinion and cause panic. Generally speaking, despite the openness of the procedure adopted by the BAPE, procedural fairness remains dubious because of Hydro-Québec=s influence on the government the BAPE=s late intervention in the process. Substantive fairness is not obvious, above all if Hydro-Québec=s concept of integrated development were to be contested. Progress must be achieved by means of clear, transparent information and a personalized approach to various clienteles. For the time being, Hydro-Québec has introduced the Électrium, an interpretation centre devoted to electricity and EMF that is designed to

168 EMF Risk Perception and Communication

initiate the public to the issue. However, other personalized means of intervention could be contemplated, such as in-home measurement of EMF to reassure anxious individuals, including measurements before and after the construction of a power transmission line.

• From the standpoint of public debate, it strikes me that the most complex question whose scope must be specified is that of false positives and false negatives and the inability of science to ever demonstrate a non-effect on health. We could refine the measurement or lengthen the period of observation. Consequently, it is inevitable that we attempt to apply to the EMF the principle of precaution put forward above all in the realms of climatic change and biodiversity.

• Beyond the theoretical debate between realists, constructivists and the proponents of a cultural approach to risk, we know that there is a considerable difference between personal risk willingly assumed and imposed risk that cannot be inferred from one to the other to legitimize an imposed risk. Sound studies written in French nonetheless reveal that individual risk is very widespread in our society and appears to reflect two different strategies, that of the limiting experience and ecstasy and that of the rite of passage or the test of bravery equivalent to the ordeal in ancient societies (Le Breton, 1995, 1996; Volant, Lévy, Jeffrey, 1996).

• Ultimately, it is in the political arena that an issue such as EMF and health unfolds. In debate such as that on EMF, it is not risk that is initially debated, nor the infinite refinement of calculation, but the quality of the social bond (Beauchamp, 1996a and 1996b). For this reason, the anticipated reforms of risk management are very important, although complex with respect to their implementation.

REFERENCES

BAPE 14, 1983. Poste Des Canton. Lignes Nicolet-Des Cantons et Des Cantons- Nouvelle Angleterre. Rapport d'enquête et d'audience publique nE 14, Québec, Bureau d'audiences publiques sur l'environnement.

BAPE 22, 1987. Projet de ligne à courant continue à "450kV, Radisson-Nicolet-Des Cantons. Rapport d'enquête et d'audience publique nE 22, Québec, Bureau d'audiences publiques sur l'environnement.

169 EMF Risk Perception and Communication

BAPE 47, 1992. Projet de la 12e ligne à 735kV. réseau d'Hydro-Québec. Rapport d'enquête et d'audience publique nE 47, Québec. Bureau d'audiences publiques sur l'environnement.

BAPE 107, 1996. Poste de ligne Duvernay-Anjou à 315kV. Rapport d'enquête et d'audience publique nE 107, Québec, Bureau d'audiences publiques sur l'environnement.

Beauchamp, A., 1996a. Gérer le risque, vaincre la peur, Montréal, Bellarmin.

Beauchamp, A., 1996b. L'électricité est-elle à risque?, Montréal, Bellarmin.

Cardinal, C., 1996. Electric and Magnetic Effects: Reflects on How to Manage the Issue - A Utility's Point of View, CIGRE - 1996 Session, 8 pages.

Hydro-Québec, 1996. Gestion des champs électriques et magnétiques, 4 volumes.

Le Breton, D., 1996 (1991). Passions du risque, Paris, Métailié.

Le Breton, D., 1995. La sociologie du risque, Paris, PUF, Que sais-je? 3016.

Levallois, P.; Lajoie, P. and Gauvin, D., 1991. Québec, Service Santé et Environnement, DSC, CHUL.

Levallois, P.; Lajoie, P. and Gauvin, D., 1994. Les risques associés aux champs électromagnétiques générés par les lignes de transport et de distribution de l'électricité. Évaluation environnementale du projet Grande-Baleine, dossier synthèse nE 9, partie 1, Bureau de soutien de l'examen public du projet Grande- Baleine.

Levallois, P. and Lajoie, P. (directors), 1998. Pollution atmosphérique et champs électromagnétiques, Québec, Presses de l'Université Laval.

Paquet, G., 1988. Les cheminements d'Hydro-Québec: émergence, épopée et désenchantement, Relations, nE 543, September 1988, pp. 202-204.

Presidential-Congressionnal Commission. On Risk Assessment And Risk Management, 1997 (2 volumes).

170 EMF Risk Perception and Communication

Volant, E.; Lévy, J. and Jeffrey, D., 1996. Les risques et la mort, Montréal, Méridien.

171 172 Principles of Risk Assessment with Application to Current EMF Risk Communication Issues

William H. Bailey Bailey Research Associates, Inc. 292 Madison Avenue, Suite 200, New York, NY 0017 Tel: (212)686-1754, FAX: (212)685-6705 Email: [email protected] and Cornell University Medical College, New York, NY

Introduction

In a previous symposium (Bailey, 1998) I drew attention to the importance of imbedding risk assessments within the risk communication process. If we acknowledge that every element of the risk assessment process has both technical and social aspects, then throughout the entire process - beginning with the way we define the issue to the way we chose to solve the issue - public involvement and communication are critical. In this 'new' concept of the risk assessment process, risk communication activities should begin before risk analysis and continue after the risk management stage (Figure 1). The first part of this paper presents a check list to rate how well a risk assessment communicates to the public and other stakeholders. To illustrate how this check list can be applied, assessments of potential health risks associated with electric and magnetic fields (EMF)42 prepared by a Working Group for the U.S. National Institute of Health Sciences (NIEHS) and the International Commission on Non-ionising Radiation Protection (ICNIRP) are rated to identify areas in which communication about EMF risks can be improved. Particular attention is given to the scientific and technical confusion surrounding the proposal by ICNIRP that EMF exposure limits for the general public be lower than for workers.

In the second part of the paper, a problem for both risk assessment and risk communication - the characterization and use of information about uncertainty - is described. The advantages of probabilistic methods for addressing this problem have been described (Bailey, 1998). Preliminary results of the use of this method in estimating the acute risks of cardiac stimulation by electric and magnetic fields are presented. In contrast to the deterministic approach used by existing guidelines to

42The use of the term EMF is usually restricted to fields with frequencies in the extremely- low-frequency (ELF) range of 0 - 300 hertz (Hz).

173 set exposure limits, this approach explicitly acknowledges and quantifies the uncertainty in estimates of 'safe' exposures to magnetic fields.

Figure 1. Illustration showing time sequence of risk assessment activities.

Aspects of Risk Assessment that Favor Good Risk Communication

Risk assessments are typically performed by experts in science and technology and their reasoning and conclusions are summarized in complex technical reports. Such reports are poor vehicles for communicating information about risk because decision makers and the public are unlikely to understand, believe, and make appropriate use of technical risk assessments. Nevertheless, it is possible to minimise such obstacles to good risk communication. Below I have summarized seven characteristics of the risk assessment process which favor good risk communication.

1. Openness to input

It is essential that the objectives of the risk assessment be responsive to the interests and concerns of the stakeholders or other audiences to whom the results of the risk assessment will be communicated. Unless the public and other stakeholders have the opportunity to identify what issues and what methods should be considered, and even which scientists might participate

174 in the risk assessment, the relevance and credibility of the risk assessment may be undermined.

2. Evaluation according to predetermined scientific criteria

Risk assessments should follow methodologies generally accepted and recognised by scientists and regulatory agencies. While all scientists are trained in scientific methods, few scientists are familiar with the formal scientific methods used to assess health risks. Guidelines for the evaluation of research to establish human health risks that are widely cited are EPA (1996) and IARC (1992). The application of these methods serves two purposes. First, it helps to ensure that the assessment is grounded upon good scientific principles. Second, it helps to ensure that the assessment is organized and carried out objectively without favoritism to any stakeholder group.

3. Comprehensive assessment

Risk assessments need to be based on all the relevant data. Many risk assessments fail to define the scope of the information to be evaluated and the method by which data were selected for consideration and weighted. When this happens, it is difficult to assess the depth and breadth of the assessment. A more vexing problem is that if the risk assessment is not totally comprehensive, or fails to establish a rational approach to the selection of data to be considered, then it may appear that the outcome has been biased by the selection of data that support a particular point of view.

4. Thorough documentation

Thorough documentation of the data, methods, and assumptions used in a risk assessment is good scientific practice. Documentation permits others to judge the comprehensiveness of the evaluation - whether there are gaps in the data or the data are not up-to-date - and provides the primary sources to determine whether the findings were fairly characterised or used in the risk assessment.

5. Transparent evaluation

175 By describing the methods by which the assessment was conducted in detail, and giving examples if necessary, the entire assessment becomes more 'available' to a wide range of experts and non-experts. While thorough documentation of the data and relevant literature consulted is essential, it is not sufficient to ensure that readers and users of the assessment can evaluate the methods and process by which the evaluation of the data was performed. From a risk communication perspective, the presentation of a transparent evaluation also demonstrates a willingness to open the risk assessment to independent scrutiny and evaluation by the broader scientific community as well as other stakeholders, and the confidence that others will understand, if not confirm, the conclusions of the assessment.

6. Acknowledgment of uncertainties

Scientists need to spend more effort in communicating the uncertainties in such a way that the confidence of the public and regulators is increased, not diminished, by the way in which issues relating to uncertainty and reliability of the assessment are addressed. While scientists are often keenly aware of all the potential limitations of the data, methods, and their interpretations and applicability, the public and regulators typically want assessments and guidance to reflect a total lack of uncertainty about any aspect of the risk being considered. The desires of the public and regulators put tremendous pressure on scientists performing assessments to ignore or to limit discussion of uncertainties. Furthermore, many scientists believe that if they were to acknowledge uncertainties, public apprehension would unnecessarily increase and their risk assessments thrown open to criticism. Such beliefs should not allow scientists to mislead stakeholders.

7. Clear and unambiguous language

The need for clear and unambiguous language is obvious. However it is often overlooked when advice is given on risk communication. A difficulty with risk assessments, as well as many other scientific communications, is that they are too technical and, if not clearly written, susceptible to misinterpretation. There are daily examples in newspapers and magazines where statements taken from interviews with scientists or from their reports are misinterpreted because of poor drafting.

176 Opportunities for Improving Risk Communication:

A Look at Recent EMF Risk Assessments

Before concerns about potential health effects of EMF became a worldwide topic of discussion there was little need for attention to the impact of EMF regulatory and risk assessment activities on the communication process. But, beginning in about 1989 in the US and 1992 in Europe, public concern became widespread. Regulators and risk assessors were unprepared to respond to this new level of concern. In response to greater concern with little new data and a lack of understanding of uncertainty, the concept of 'prudent avoidance' was seized upon as an interim policy option and cornerstone for communication. Although initially defined as the exercising of fiscal prudence by individuals in dealing with speculative or unknown health risks (Morgan, 1989), it soon was popularized to refer to precautionary actions taken to mitigate likely, but not totally definitive health risks. With the advent of many new epidemiology and laboratory research studies, there is now a firmer basis for risk assessment and the opportunity to examine the need for new policies and guidelines regarding EMF exposures.

NIEHS Working Group and ICNIRP Reports

Recently, new risk assessments have been offered to update legislative actions and exposure guidelines. These risk assessments have been published by a Working Group organized by the National Institute of Environmental Health Sciences (NIEHS, 1998) and the International Commission on Non-Ionising Radiation Protection (ICNIRP, 1998). The goal of the risk assessment prepared by the Working Group is to identify the level of risk, if any, that is associated with electric and magnetic field exposure and develop a basis for the US Congress to propose legislation to manage or limit exposures. The foci of concern in this risk assessment are adverse health effects, cancer being the most important, alleged to be caused or exacerbated by EMF exposure. The risk assessment prepared by ICNIRP addressed these same concerns but tersely concluded ". . . that the results from the epidemiological research on EMF field exposure and cancer, including childhood leukaemia, are not strong enough . . . to form a scientific basis for setting exposure guidelines." ICNIRP then addressed the issue of short term effects of fields at frequencies less than 1 kHz related to induced currents, surface electric fields, and contact currents. The ICNIRP risk assessment serves as the basis for recommendations to national authorities for exposure limits of workers and of the general population.

177 Ratings of Effective Risk Communication

Stakeholders need more and better communication about EMF and health research. It is therefore important to assess how much progress has been made in using risk assessments as a vehicle for improved communication. Using the seven characteristics of effective risk assessments described earlier as a nominal set of evaluation criteria, I rated the ICNIRP and NIEHS risk assessments on an ordinal scale from 1-10 with one representing the most effective and 10 the least effective in communicating risk (Table 1).

Table 1 Opportunities for Improving Risk Communication Risk Assessment Characteristics Ratings Favoring Effective Risk 1998 ICNIRP NIEHS WG Communication

Open to input 7 1

Predetermined criteria 3 2

Comprehensive assessment 6 4

Thorough documentation 8 3

Transparent evaluation 8 2

Characterisation of uncertainties 9 9

Clear and unambiguous language 7 5 Rating Scale 1 = Small opportunity for improvement 10 = Great opportunity for improvement

For those who are familiar with these two risk assessments, it is obvious that the differences in their ratings vis à vis risk communications largely reflect different organizational goals and policies and not differences in their fundamental scientific

178 quality. With an awareness of the need for risk communication, future EMF risk assessments can be designed to support this function as well as meet scientific goals. This may be difficult for organizations that set guidelines because of insufficient budgets and staff resources, time, and planning. More important, risk communication may not even be a part of their mission. In general, insufficient planning for risk communication is common where risk communication has been selected by default as the 'last task' of a risk assessment.

The low ratings given to both risk assessments for characterization of uncertainty reflect inadequacies in describing uncertainties in the underlying scientific data and in providing a method for addressing these uncertainties. A later section of this paper discusses how the uncertainties regarding effects of tissue stimulation by induced currents - the basis for all exposure guidelines to date - can be characterised by probabilistic methods.

Communication About EMF: Possible Occupational Risks vs Phantom Risks to the General Public

There is sufficient theoretical and experimental basis to be concerned about possible adverse effects caused by the induction of voltages and currents in the body by intense alternating electric and magnetic fields (Bailey et al, 1997). Yet, guideline-setting organizations (including ICNIRP) and regulatory organizations, e.g., the Bureau of Radiological Devices, U.S. Food and Drug Administration have not cited as evidence a single safety incident in which an acute induction effect has occurred in occupational or other environments. Nevertheless, the possibility remains that some exposure scenario in the workplace, perhaps involving some unique technologies, conceivably could present a direct or indirect hazard to workers. Hence the need exists for guidelines like those proposed by ICNIRP and other organizations to limit the intensity of occupational exposures. ICNIRP specifies that an adequate level of protection can be achieved by setting occupational exposure levels 10-times lower than thresholds for some biological effects of concern.

ICNIRP43 has also recommended that the exposures of the general public to electric and magnetic fields in the ELF range be limited to much lower levels than those recommended for workers. From a risk assessment and risk communication perspective it is important to understand the rationale used to choose a biological

43The ICNIRP and the interim CENELEC (1995) guidelines are similar so that comments about ICNIRP often apply to CENELEC as well.

179 'dose' limit for the general population that is 5-times lower than that set for workers. By recommending limits for the general public, ICNIRP assumes a responsibility to succinctly communicate the basis for its recommendations. Unfortunately, no explanation supported by data or analysis is presented. All that ICNIRP considers is "the general public . . . may include particularly susceptible groups or individuals" (p. 508) and that there is some uncertainty regarding the appropriate safety factor because of "a lack of knowledge regarding the appropriate dosimetry." (p. 508). These statements alone without supporting data and analysis are the sole justification for requiring greater protection for members of the general public than for adult workers. The former may not be informed, trained, or expected to take precautions, but this may be unimportant if exposures below occupational exposure limits are truly without adverse effect. In his editorial for the recent special issue of Radiation Protection Dosimetry on non-ionising radiation, John Dennis laments that "[t]he present choice of 5 for this [safety] factor by ICNIRP appears to be due more to the number of fingers on the human hand than any rational evaluation. . . .[t]he justification advanced . . . [is] subjectively appealing, but hardly stands up to any logical analysis." (Dennis, 1997).

Another explanation for the recommendation of a large safety factor (10 for occupational x 5 = 50) for public exposure could include an attempt by ICNIRP to respond to concerns of the public about cancer from long-term, low level exposures. ICNIRP may have felt that the recommendation of a larger safety factor, even based on short-term effects, not cancer, would respond to, and calm, public fears and perceptions without committing to a guideline that could not be justified based on the existing epidemiology and laboratory data regarding cancer.44 Or, it might be that the purpose was to provide an additional safety factor to protect against unknown effects of long-term exposure. If so, ICNIRP should have publically discussed this approach in the interest of fully communicating the basis for its recommendation of a public exposure guideline. Until ICNIRP provides further explanation of the biological data and rationale underlying the application of the safety factor for public exposure, such speculative explanations cannot be discounted.

For electric fields, ICNIRP appears to have intended recommended exposure limits for the general public to meet two perceived needs: 1) to minimise internal induced current densities; and 2) to minimise perception of electric fields at the surface of the body and perception of indirect contact currents. The guideline

44A close reading of the rationale given by DG V (1997) for limits on public exposures to EMF would appear to follow this interpretation.

180 attempts to simultaneously address both goals yet does not explain how this affects the selection of exposure limits.

The basic restrictions given in the ICNIRP guideline are specified in terms of induced current density. Studies in the literature are cited to show that current densities above 100 mA/m2 might have adverse stimulating effects on the nervous system. No rationale for exposure guidelines other than the prevention of such effects is given. Because ICNIRP points to no evidence for cumulative effects of exposure, one must therefore conclude that for these phenomena of concern, exposures less than the occupational exposure criterion, 10 mA/m2, would not adversely affect either workers or the general public. Hence, it is not at all clear how ICNIRP can justify a lower exposure limit for the general public than for workers based on potential risks of current induction alone. Indeed, the Electromagnetic Field Project of the World Health Organization states that "there is no need for any specific protective measures [from ELF fields] for members of the general public." (WHO, 1998).

For electric fields, ICNIRP justifies basic restrictions and reference levels to limit exposures of the general public to levels below those recommended for occupational environments on the prevention of surface field effects. Although painful shocks might be experienced in electric fields greater than 10 kV/m when grasping large conducting objects, sensations experienced at lower field levels in touching smaller objects are merely annoying (CRP, 1997). Hence, the electric field limitation for the general public is designed to prevent perception, not "detectable impairment of the health of the individual." The meaning of this position is further clarified by ICNIRP's Response to Questions and Comments on ICNIRP Guidelines (Matthes, 1998) which states, "[t]he reference levels for electric fields at power frequencies were set to limit indirect effects of contact with electrical conductors in the field."

Although the relationship between the limits for electric field exposure and contact currents are not discussed in the guideline, there is obviously considerable overlap in their goals. The rationale to limit induced current densities in the general public to levels one-fifth of those recommended for workers is inconsistent with ICNIRP's assertion that "[t]he [occupational reference level] value of 10 kV m-1 for a 50-Hz or 8.3 kV m-1 for a 60-Hz occupational exposure includes a sufficient safety margin to prevent stimulation effects from contact currents under all possible conditions." (p. 510). The WHO EMF Information Project also states, "the effects of exposures of up to 20 kV/m are few and innocuous" (WHO, 1998). This raises

181 the question why reference limits on electric fields are needed to prevent field perception by the public if ICNIRP has concluded there are no significant risks for this exposure and has recommended separate reference limits for contact currents.

The above discussion highlights the need for ICNIRP to initiate a more extended discussion of its guideline setting process so that important questions regarding the formulation and communication of guidelines for public exposure to ELF fields can be resolved.

Progress in Characterizing Risks of EMF Induction

Communications about risks can be improved by conveying the type and magnitude of the uncertainties associated with an evaluation or risk estimate. Yet all too often scientists are unfamiliar with the tools for addressing and communicating uncertainties or shy from discussing these uncertainties lest lay persons 'misinterpret' the discussion. If there is concern that the mention of uncertainty in a risk assessment may cause the public to either lose faith in the credibility of the assessment or overestimate the magnitude of potential risks, it is the responsibility of scientists to do a better job in presenting and communicating the uncertainties. In the previous symposium I have summarized the advantages of a probabilistic approach to risk assessment, including communication and guideline setting (Bailey, 1998). The advantages of this approach also apply to assessing the potential risks of EMF induction (Bailey, 1997). To update our progress in quantifying relevant uncertainties with probabilistic methods, I have performed preliminary analyses to illustrate how a probabilistic method approach is used to calculate the likelihood of acute risks of cardiac stimulation from exposure to power frequency magnetic fields. The details of the method and the results are described elsewhere (Bailey et al, 1998).

Probabilistic Method for Modeling 'Safe' Magnetic field Exposures

A model was constructed to predict a level of whole-body magnetic-field exposure below which no cardiac stimulation (ncs) is expected (Bncs). The model demonstrates how the estimation of 'safe' exposure levels is affected by uncertainties in critical input variables. Stimulation of the heart by ELF magnetic fields was chosen as a critical effect because it is a clearly identifiable, potentially adverse effect that is typically considered in setting guidelines for occupational exposures to magnetic fields. A probabilistic model offers advantages: it is possible to estimate the likelihood of

182 effects at specific exposure levels and with sensitivity analyses it is possible to determine priorities for future research.

The form of the model was chosen to reflect the dosimetric approaches used by existing guidelines (Bailey et al, 1997).

Bncs < (CST) / (S * F * C) (1)

B is the root-mean-square (rms) magnetic-field level in tesla (T), CST is the cardiac-stimulation-threshold (CST) current density in amperes per meter-square (A m-2), S is the shape factor in meters, F is frequency in hertz (Hz), and C is conductivity in siemens per meter (S m-1). The CST is the induced current density required to stimulate contractions of the heart. For this study, probability distributions were developed to characterise the CST, S, and C input variables. The input distributions were developed based on empirical data and/or assumed probability distributions. The values chosen were meant to be consistent with the range of values typically referenced in ELF guidelines. Initially, frequency was assumed to be fixed at 60 Hz; in later studies, not discussed here, a distribution representing 60 Hz and higher harmonics was also derived.

Distribution of Bncs were computed for two scenarios by selecting single values from the CST, frequency, shape factor, and conductivity distributions by Latin Hypercube sampling and calculating a value for Bncs using the model described by expression (1). The two scenarios differed in the dosimetric model assumed; one used the circular loop model described by IRPA/INIRC (1990), the other used the ellipsoid model described by ACGIH (1996). Each Bncs value represents a calculated threshold for a combination of circumstances such that whole-body magnetic field exposures lower than Bncs are without risk of cardiac stimulation. The input distributions were sampled and Bncs values calculated on 5000 trials. Distributions of the Bncs values obtained reflect the effect of the combined uncertainty of all the input variables on the predicted value of Bncs.

The frequency distributions of calculated Bncs values assuming the circular- loop dosimetric model used by IRPA/INIRC (1990) and ICNIRP (1998) and the ellipsoid dosimetric model used by ACGIH (1996) are shown in Figure 2 for a 60- Hz magnetic-field source. For convenience, the data are presented as complementary (i.e., reversed) cumulative probability distributions. The graph shows the relationship between magnetic-flux density and the probability that

183 stimulation would not occur i.e. the probability of obtaining a value less than Bncs calculated by the model.

Insights from Probabilistic Modeling

The wide range of simulated Bncs magnetic-field values indicates that there is considerable uncertainty about what constitutes a safe level of exposure predicted by either model.

100%

90%

80%

70% 60Hz-circular 60% 60Hz-ellipsoid 50%

40% Percentile

30%

20%

10%

0% 0.001 0.01 0.1 1 10 100

Calculated Value of Bncs (T)

Figure 2 Distributions of Bncs for 60-Hz magnetic-field exposures based on circular-loop and ellipsoidal dosimetry models.

However, all of the uncertainty relates to exposures greater than about 0.01 T. About a tenfold increase in exposure level to 0.1 T would be required before the risk of stimulation occurring would be about 5 %. These results confirm that the major occupational-exposure guidelines are very conservative with respect to risks of cardiac stimulation. The minimum Bncs values computed here are roughly ten times greater than the workday magnetic field exposure values recommended by ACGIH (0.001 T) or ICNIRP (0.004 T) at 60 Hz. The lowest 5% Bncs value calculated for the circular-loop model is about 50-times the IRPA/INIRC (1990) and ICNIRP (1998) occupational exposure limits; the lowest 5% Bncs value calculated for an ellipsoid model is more than 100-times the ACGIH (1998) occupational

184 exposure limit. The probabilistic analysis thus supports the degree of conservatism that was suggested from a cruder analysis (Bailey et al. 1997).

The differences between the Bncs estimates obtained with the ellipsoid model and those obtained with the circular-loop model indicate the uncertainty associated with the choice of dosimetry model. Guidelines frequently cite stimulation of peripheral nerves as also of potential concern. The implications of this analysis are relevant to this effect as well because the threshold for stimulation of peripheral nerves in humans may be only slightly below that for cardiac stimulation (Reilly, 1992) and the relevant variables affecting the stimulation of nerves are similar to those affecting the heart.

To characterise the relative importance of each input variable to the overall uncertainty in Bncs, probabilistic sensitivity analyses were performed. The magnitude of the Spearman rank order correlation (r) between the values of the input distributions and the output Bncs distribution provides a relative estimate of the contribution of each input variable to the uncertainty in the Bncs distribution based on their entire range of values. The results of the sensitivity analysis for the ellipsoidal model are shown in Figure 3. None of the input variables are strongly associated with the distribution of calculated magnetic-field levels. However, the correlation analysis indicates that, of the variables considered, the CST is the variable that most strongly affects the uncertainty in the predicted values of Bncs. The results of sensitivity analyses for the circular-loop dosimetry model are similar (data not shown).

The most important insight provided by the analysis that the greatest source of uncertainty as to what is a "safe" level of exposure to magnetic fields is knowledge as to the CST. Although the ranking of the CST was just marginally higher than the other input variables to the model, its significance is that it appears at all as a major contributor to the uncertainty in Bncs. In recent decades, guidelines have given very little attention to the need to verify and characterise biological thresholds of tissues to stimulation by induced currents. Much of the biological data regarding stimulation thresholds that have been considered in the setting of guidelines for ELF field exposure are out-of-date, inconsistent, and characterised by considerable uncertainty (Bailey et al. 1997). Despite interest at this time in effects of induced currents and the development of new exposure guidelines, the literature has not to this point addressed the critical importance of defining biological thresholds in the determination of safe exposures.

185 C orrelation C oefficients (r)

C a rd ia c S tim u la tio n T h r e s h o ld

0.62

C onductivity

-0 .5 0

S hape Factor

-0 .2 5

-1.00 -0.50 0.00 0.50 1.00 C oefficient Value

Figure 3 Probabilistic sensitivity analysis for ellipsoid dosimetric model showing Spearman rank order correlations between cardiac stimulation threshold, conductivity and shape factor values, and calculated Bncs values

Summary

A new concept for risk assessment is that risk communication activities anticipate and continue throughout the assessment process. This means that the risk assessment process and its findings be fully transparent and available to scientists and the public. This paper discusses seven characteristics of science-based assessments that can improve the effectiveness of risk communication to the public and decision makers. I rated the EMF risk assessments published by NIEHS and ICNIRP on these characteristics to illustrate opportunities for improved risk communication. The observation that my subjective ratings are largely determined by the purposes and goals of these risk assessments should foster a closer look at structuring future risk assessments to meet both risk communication and scientific goals.

One particular recommendation of the ICNIRP risk assessment was a guideline to limit exposures of the general public to electric and magnetic fields at frequencies less than 1 kHz. As discussed here, the scientific basis for this

186 recommendation needs to be examined more carefully. It is also important that the purpose and rationale be clarified so that the general public will not mistakenly perceive that exposures to EMF at levels commonly encountered in the environment pose any likely health risk.

Both the NIEHS and ICNIRP EMF risk assessments were given low ratings for the characterization of uncertainties. One way to address uncertainty is to develop quantitative characterizations of the potential risks of EMF for various exposure scenarios using probabilistic models. The method is also applicable to describing a realistic range of risks associated with specific exposure scenarios in quantitative terms.

References

American Conference of Governmental Industrial Hygienists (ACGIH). Documentation of the threshold limit values for physical agents in the work environment. Cincinnati, OH: ACGIH; 1996.

American Conference of Governmental Industrial Hygienists (ACGIH). Threshold limit values for chemical substances and physical agents. Cincinnati, OH: ACGIH; 1998.

Bailey, WH. Probabilistic approach to deriving risk-based exposure guidelines: application to extremely low frequency magnetic fields. Radiation Protection Dosimetry 72:327-336; 1997.

Bailey, WH. Principles of risk assessment and their limitations. In Risk Perception, Risk Communication and its Application to EMF Exposure. Matthes, R; Bernhardt, JH; Repacholi, MH. (eds). Oberschleißheim, Germany: International Commission on Non-Ionising Radiation Protection, 1998.

Bailey, WH; Su, SH; Bracken, TD. Probabilistic approach to ranking sources of uncertainty in ELF magnetic field exposure limits. Submitted for publication, 1998.

Bailey, WH; Su, SH; Bracken, TD; Kavet, R. Summary and evaluation of guidelines for occupational exposure to power frequency electric and magnetic fields. Health Physics. 73:433-453; 1997.

187 Barnes, D.G.; Dourson, M.. Reference dose (rfd): description and use in health risk assessments. Regulatory Toxicology and Pharmacology. 8:471-486; 1988.

Comité Européen de Normalisation Electrotechnique (CENELEC). Human exposure to electromagnetic fields, low frequency (0 Hz to 10 k Hz). CENELEC Standard ENV 50166-1; January 1995.

Commission on Radiation Protection (CRP). Protection against low-frequency electric and magnetic fields in energy supply and use. In Berichte der Strahlenschutzkommission des Bundesministeriums für Umwelt, Naturschutz und Reaktorsicherheit. Heft 7. Stuttgart: Fischer, 1997.

Directorate-General V Employment, Industrial Relations and Social Affairs of the European Commission (DG V). Draft Commission proposal for a Council recommendation on limits for exposure of the general public to electromagnetic fields, November 27, 1997.

Dennis, J. Editorial: Non-ionising issues. Radiation Protection Dosimetry. 72: 3-4, 1997.

Durney, C.H.; Massoudi, H.; Iskander, M.F. Radiofrequency Radiation Dosimetry Handbook. United States Air Force School of Aerospace Medicine. USAFSAM- TR-85-73; October 1986.

International Agency for Research on Cancer (IARC). Working Group on the Evaluation of Carcinogenic Risks to Humans. IARC monographs on the evaluation of carcinogenic risks to humans. Volume 55: Solar and ultraviolet radiation. Views of the IARC Working Group meeting in Lyon, France, February 11-18, 1992. United Kingdom: International Agency for Research on Cancer (IARC). Distributed by the Secretariat of the World Health Organization (WHO), 1992.

International Radiation Protection Association/International Non-Ionising Radiation Committee (IRPA/INIRC). Interim guidelines on limits of exposure to 50/60- Hz electric and magnetic fields. Health Physics. 58:113-122; 1990.

International Commission on Non-Ionising Radiation Protection (ICNIRP). Guidelines for limiting exposure to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz); Health Physics. 74: 494-522; 1998.

188 Matthes, R. Response to questions and comments on ICNIRP. Health Physics. 75:438-439; 1998.

Morgan, G. Electric and Magnetic Fields from 60 Hertz Electric Power: What Do We Know about Possible Health Risks? Pittsburgh: Carnegie Mellon University, 1989.

National Institute of Environmental Health Sciences (NIEHS). Assessment of Health Effects from Exposure to Power-Line Frequency Electric and Magnetic Fields. Working Group Report. NIH Publication No. 98-3981. Research Triangle Park, NC: National Institute of Environmental Health Sciences of the U.S. National Institutes of Health, 1998a.

Reilly, J.P. Electrical stimulation and electropathology. Cambridge, England: Cambridge University Press; 1992.

United States Environmental Protection Agency (USEPA). Proposed guidelines for carcinogen risk assessment. Federal Register, 61:17960-18011; 1996.

World Health Organization. WHO Launches and Initiative to Harmonize Electromagnetic Field Standards Worldwide. Press Release, 17 November, 1998.

Acknowledgments

The work described in this paper was supported by research contracts from EPRI (WO2966-14) and Electricité de France-Gaz de France. The comments and suggestions of T. Dan Bracken are greatly appreciated.

189

How Dangerous Is It Really? Some Approaches for Inferring Risk

Daniel Wartenberg Environmental and Occupational Health Sciences Institute UMDNJ RW Johnson Medical School 681 Frelinghuysen Road, P.O. Box 1179 Piscataway, N.J. 08855-1179, USA Tel: +1 732 445 0197, Fax: +1 732 445 0784 E-mail: [email protected]

The assessment of the possible human health effects of exposure to electric and magnetic fields remains a controversial issue. Although two recent blue ribbon panels have concluded that there is limited evidence of an association,16 many remain skeptical and some disagree strongly.17 I will address this issue in two parts. First, one must determine whether an agent causes disease. Second, if causation is shown (or accepted at least for argument s sake), one can infer the magnitude of this risk. In this presentation, I briefly review the criteria for causation and then consider three different approaches for assessing possible human health risk: (1) weight of evidence review; (2) meta-analysis; and, (3) quantitative risk assessment. I apply each to the magnetic field data.

Causation

In modern epidemiology, assessments of causation most often refer back to Sir Bradford Hill s Presidential Address to the Royal Society of Medicine in which he outlines nine aspects of an association that one ought to consider, before deciding that the most likely interpretation of it is causation. 8 There are two statements in Hill s paper that I highlight as these issues are often overlooked. First, the list Hill presents is a set issues to consider in evaluating the likelihood of causation, not a list of strict criteria. In his own words, I do not believe we can usefully lay down some hard-and-fast rules of evidence that must [his emphasis] be obeyed before we accept cause and effect none can be required as a sine qua non. Rather, one must undergo a careful assessment of the available data and studies in reaching a conclusion about a body of work. And yet, each body of information differs, and one may have to make different judgements about data for each agent and/or situation.

191 To reach a conclusion, Hill suggests that in light of the set of available observations one should consider whether there is any equally likely or more likely explanation than cause and effect. That is, one should review all possible explanations for a given set of observations and accept causation as an explanation only if there is no other explanation that is at least as well supported by the data.

The second often overlooked point Hill raises is that one should consider the aspects of causation only after one has observed, an association between two variables, perfectly clear-cut and beyond what we would attribute to the play of chance. In other words, he suggests ruling out random error prior to the consideration of causation, as one might do using meta-analysis. Results that are equivocal from a statistical perspective do not merit consideration as causal and thus should not be considered until the data more clearly show an unusual set of results or circumstance.

The nine items Sir Hill lists for consideration are: strength, consistency, specificity, temporality, biological gradient, plausibility, coherence, experiment, and analogy. One must use judgement considering each of these, and integrating them into a holistic evaluation. In concluding, Hill notes that tests of significance cannot answer the question of causation. As Sander Greenland paraphrases him, one must be weary of equating statistical significance with scientific significance.6 That being said, we consider three approaches for assessing causation and inferring risk.

Weight of Evidence Review

Weight of evidence review is a careful and systematic evaluation of a body of literature. Studies ranging from in vitro to in vivo to epidemiology are considered in reaching an overall assessment of the possible carcinogenicity of a substance. While literature searches are often used to identify appropriate studies, generally the specifics of the literature search, the databases accessed and the criteria for inclusion or exclusion of specific studies are not stated explicitly. Then, the review of each study is also subjective, based on the reviewer's best judgement of the quality, relevance and importance. Typically, inference is guided by consideration of various criteria for causality.8 There is concern that investigators may idiosyncratically focus on particular studies or specific issues within studies that are not representative of the whole body of the literature.

Agencies have developed specific criteria for undertaking such evaluations. For example, the International Agency for Research on Cancer has a long-standing

192 program for the review of agents that are suspected of being human carcinogens.9 As part of the review process, members of the panel conducting the review determine which studies are of sufficient quality and relevance to be included in their deliberations. While there are explicitly stated criteria for inclusion/exclusion criteria, some question the objectivity of this approach and its application to a set of studies.

Similarly, the United States Environmental Protection Agency, through its risk assessment program, reviews the possible human carcinogenicity of suspected agents.20,21 Again, a careful but subjective review is conducted. Some interested parties often challenge the objectivity of the process.

Both of these programs consider the full range of studies, from laboratory to human, for evaluation. They also propose classification schemes to summarize findings, addressing both the adequacy of the data and the likelihood of carcinogenicity. Paraphrasing, the categories differentiate agents not likely to be carcinogenic, from those possibly carcinogenic, to those likely to be carcinogenic to people. Often holding meetings of panels of experts, these agencies ask scientists for their best judgement about the carcinogenicity of specific agents. While criteria for evaluation are set out, the final decisions are ones of interpretation of a body of literature as a whole.

Meta-Analysis

Meta-analysis is the systematic review of a body of literature and the application of statistical methods to summarize (i.e., average) the quantitative results of individual studies, in the hopes of identifying consistent patterns and sources of disagreement among those results.2,7 To conduct a meta-analysis, all epidemiologic studies are retrieved systematically and reviewed on the basis of a consistent set of criteria. This review also can be achieved by a more traditional literature review, but those efforts typically are less systematic and less comprehensive. Some meta- analyses evaluate the completeness of the study selection by conducting a test for publication bias.3,19 These tests assess whether studies with large effect sizes have greater precision, as would be expected, or how many null studies would it take to reduce a statistically significant average effect to a non-statistically significant average effect. If the set of identified studies is viewed as appropriate, meta- analysis can be used to assess the consistency, comparability and heterogeneity of results of each of the identified studies. Finally, if the studies are sufficiently homogeneous, meta-analytic tools can be used to estimate average study relative

193 risks. If heterogeneous, meta-analysis can be used to try to explain the differences in study results. Sometimes, the influence of individual studies on the average effect size is assessed by excluding each study and determining how much the average effect would change.18 This is considered particularly for index studies, those which first generated concern on an issue, or for studies of poor quality.

Quantitative Risk Assessment

The primary goal of quantitative risk assessment is to estimate the hazard for an exposure or situation that for some reason cannot readily be measured directly.15,23 Where possible, epidemiologic assessment typically provides a more valid and reliable estimate of the human health hazard than does risk assessment. But often, we cannot conduct or do not have time to wait for the completion of an appropriate epidemiologic study. For example, one may wish to estimate the hazard that will result from some activity that is proposed but has not occurred (e.g., application of a particular pesticide to the food supply or the construction of a new incinerator) or one in which the complexities of the true situation make it too difficult to study (e.g., wide variations in the diet of a small population or the impact of emissions from an incinerator in an urban setting). In such situations, quantitative risk assessment offers a more tractable approach to infer the magnitude of the potential hazard in a timely manner.

There are four main stages of risk assessment: hazard identification, exposure assessment, dose-response analysis and risk characterization. Hazard identification is the identification of all situations or substances that can, under any circumstances, pose a risk to human health, and all adverse health effects that could possibly result. It is meant to include all hazards regardless of the size or amount, and all possible health endpoints regardless of the likelihood of detectable response. If a hazard cannot be identified, many believe the risk assessment should be discontinued.

Exposure assessment is the estimation for each situation or hazard listed in the Hazard Identification of the amount of exposure to the hazard that a typical person is likely to encounter. This must include a characterization of the source(s) of exposure and the intersection of a person s activity pattern with the hazard due to the source.

Dose-response analysis is the stage in which one determines generically the amount of exposure which causes harm. For carcinogenicity assessed in animals,

194 this is typically called the "cancer potency." For adverse outcomes assessed in epidemiologic studies, the measure of effect is usually called the relative risk. Typically, risk assessors fit a mathematical equation to the data to describe how the risk of disease increases with the amount of a substance a person is exposed to.

Risk characterization is the stage in which the information from the three other stages is combined into a single overall estimate of risk. That is, for the hazard, the exposure information is combined with the dose-response information to predict a risk of an adverse effect for some exposed at the specified level. Risk can be reported for an individual, or for a population as a whole.

Application to Magnetic Field Data

By comparing the application three approaches presented above to magnetic field data, one can begin to appreciate where and why some of the disparities in scientific assessments have arisen.

Weight of Evidence

The weight of evidence approach has been applied several times to the EMF issue, by both individuals and expert panels. There have been over two dozen expert panel reviews of the EMF issue, far too many to review here. The two most recent reviews were conducted in the United States.16,17 The first,16 under the sponsorship of the National Academy of Sciences, utilized their own process for evaluation (Table 1). First, a series of workshops were held at which the group was briefed on the state of the research. Then, members reviewed papers in their own fields of expertise and wrote summaries. These summaries were then considered and commented on by the whole group until consensus was reached and then formed the basis for the final report. Their bottom line was that while, no conclusive and consistent evidence shows that exposures to residential electric and magnetic fields produce cancer, they also asserted that, an association between residential wiring configurations and childhood leukemia persists in multiple studies Interpretations of these statements varied.25

The second panel,17 convened by the National Institute of Environmental Health Sciences (NIEHS), was instructed to follow the procedures developed by the International Agency for Research on Cancer (Table 1).9 After a series of topical workshops during the months prior to the panel meeting, commissioned draft chapters of the final report were written and circulated among those invited to attend

195 the Working Group meeting. At the Working Group meeting, these chapters were reviewed, edited and rewritten until the subgroups approved them. Then, the Working Group as a whole considered each part of the report, making comments, editing and eventually voting on the acceptance and specific conclusions of each chapter and the report as a whole. They concluded that, ELF EMF are possibly carcinogenic to humans (Group 2B). This was based principally on, the results of studies on childhood leukemia in residential environments and on CLL [chronic lymphocytic leukemia] in adults in occupational settings. In addition, the participants stated that the in vitro and mechanistic data provide weak support based on studies at very high levels of exposures (>100 µT).

Table 1: Weight of Evidence

NRC16 NIEHS17 Foster5 Emphasis Residential Residential and In vivo and in epidemiology occupational vitro studies epidemiology Observed effects Childhood Childhood leukemia, None driving concerns leukemia adult chronic lymphocytic leukemia Causation No consistent and Possible human Not a probable conclusive carcinogen carcinogen evidence Quantitative Not conducted Conducted Not conducted Risk Assessment

One recently published article using weight of evidence criteria reached somewhat different conclusions (Table 1). 5 After reviewing the literature, they placed greater emphasis on in vitro and in vivo studies than the expert panels, pointed out the lack of genotoxicity of EMF, the lack of plausible biological mechanism and the apparent inconsistencies between epidemiology study results, and concluded that the evidence in support of links between the [electromagnetic] fields is weak and inconsistent. The assert that, the issue is not whether fields could possibly be related to human cancer but whether they probably are [original emphasis]. The latter sentence is interesting in that it reframes the question into one that requires a much higher and more consistent standard of evidence.

Meta-Analysis

196 Meta-analysis is conducted only if you believe that the epidemiology is reliable and of sufficient quality for further consideration. Some EMF investigators believe it is not. Others do and several meta-analyses have been conducted for both the residential1,12-14,16,22,24,26 and occupational10,11 exposures.

The most recent residential exposure meta-analysis,24 conducted as part of the NIEHS Working Group review noted above, extends previous work by providing:

1. analyses including several more recent studies;

2. a more careful and detailed review of the exposure metrics motivating and justifying particular analyses;

3. estimates for fixed effects, random effects, heterogeneity, and sample size needed to negate the observed results, for a variety of exposure classifications;

4. analyses stratified by various study characteristics to provide preliminary evaluation of possible confounding and effect modification;

5. analyses for publication bias, including the fail-safe N and funnel plots;

6. dose-response meta-analysis including heterogeneity assessment.

Results are summarized in Table 2. First and foremost, it is important to review the analyses for heterogeneous studies. The study showed intermittent heterogeneity, although sometimes it was substantial. When present (e.g., wire codes scored by 24-hour measurements), summary estimates of effect can be misleading. Stratification to identify particular sources of heterogeneity did not provide specific insights. Overall, summary relative risks ranged between 1 and 2, and these were not subject to substantial influence by individual studies. There was only limited evidence of publication bias. A large number of not statistically significant unpublished studies would have to exist for the observed results to be due to random variation (according to the Fail-Safe N), and an extremely large, negative study would be required to reverse these observations (the number needed calculation). Based on these findings, the association between wire codes and childhood leukemia is most striking, and the association between measured fields and childhood leukemia also difficult to explain. While confounding and bias

197 cannot be ruled out as possible explanations, several investigations into these factors have failed to identify plausible factors responsible for the observed effects.

Table 2: Meta-Analysis

Criterion Index Measured/Calculated Fields Proximity to Source Dichotomy ContinuousDichotomy Continuous Spot Calculated Wire Wire Measure- Fields Codes Codes ments scored by scored by spot 24-hour

Strength Summary 1.4 1.1 1.2 1.4 2.7 1.6 RRA (1.0-2.0) (0.9-1.3) (0.9-1.5) (1.1-1.8) (0.8-8.7) (0.5-4.6) Consis- % positive 80% 75% 75% 73% 100% 50% tency studies (number of (10) (4) (4) (11) (2) (2) studies) Hetero- 0.2 0.3 0.2 0.1 0.1 0.02 geneity Publica- Fail-safe 7----30---- tion NC BiasB Subjects >6000 -- -- >3400 -- -- needed Influence Hetero- 0.11-0.50 -- -- 0.04-0.20 -- -- Analysis geneity Relative 1.2-1.6 -- -- 1.3-1.5 -- -- Risk

ARandom Effects Model (Reference 4) BCalculated only for dichotomous exposure categorizations CCalculated only if result is statistically significant (References 3,19)

Quantitative Risk Assessment

The first stage of quantitative risk assessment is the identification of likely hazards and outcomes. If present, one proceeds; if not, one does not conduct a risk assessment. The National Academy of Sciences Committee did not believe that there was sufficient evidence of causation to conduct a risk assessment of EMF exposure.16 The NIEHS Working Group, however, concluded that EMF was a possible human carcinogen (Group 2B) and therefore believes that there is sufficient evidence to conduct the assessment.17 The outcomes they identify are childhood leukemia for residential exposures and adult chronic lymphocytic leukemia for occupational exposures. NIEHS staff are currently conducting such an assessment. Below, I provide a preliminary and cursory quantitative risk assessment of

198 residential exposure to EMF based on the data from the NIEHS Working Group Report.17 Results are summarized in Table 3.

Having identified both the hazard (residential EMF exposure) and the outcome (childhood leukemia), we next have to quantify exposure. Using exposure data developed from surveys of homes throughout the US, we have both distributions of wire codes and spot measured magnetic fields.27,28 For wire codes, it was reported that 28% of homes have ordinary high (OHCC) or very high (VHCC) current configurations. For spot measurements, the data were reported to follow approximately a lognormal distribution with a mean of 0.09 µT and a standard deviation of 2.2 µT. Using the relative risks of 1.4 for OHCC or higher wire codes and 1.1 per 0.1 µT for spot measured magnetic fields, as reported in the NIEHS meta-analysis,24 and the reported annual 2,200 cases of leukemia cases to children under 15 years of age (source: Leukemia Society of America), one can calculate the number and proportion of cases attributable (PAR) to EMF exposure each year. Based on the wire code data, we predict about 175 cases, or 8%, attributable to EMF exposure. Based on the spot measurement data, we predict about 240 cases, or 11%, attributable to EMF exposure. From a policy perspective, these are substantial numbers, if somewhat uncertain.

Table 3: Quantitative Risk Assessment

Stage of Risk Assessment Wire Codes Measurements Hazard ID Group 2B carcinogen Group 2B carcinogen Exposure Assessment 28% > OHCCA Lognormal (0.09,2.2)B Dose ResponseC RR=1.4 RR=1.1 per 0.1 µT Risk PAR 8% 11% Characterization # cases ~175 ~240

AZaffanella27 BZafanella28 CWartenberg24

Discussion

To summarize the disparate viewpoints on the possible human health risk from exposure to magnetic fields, one can review Hill s aspects of causation in assessing the plausibility of a hazard (Table 4), and one can consider the risk assessment to estimate (albeit crudely) the magnitude of that possible risk (Table 3).

199 First, one sees a fairly weak but fairly consistent effect in the epidemiologic studies reported. While the result is not highly specific, there is evidence of temporality and biological gradient. Relatively newly developed models do exist which may explain adverse effects at high dose in in vitro studies, levels which border on rare occupational levels, although models have yet to be developed that are relevant at doses equivalent to residential exposures. Thus, there is some support for causality from the perspective of Hill s approach, but more data are needed to further clarify this issue.

Table 4: Hill s Aspects of Causation Applied to the Magnetic Field Issue

Strength RR=1.1-2.7 Consistency Heterogeneity varies but appears moderate with exceptions Specificity Other causes are known but explain only a small proportion of total cases Temporality Yes Biological Some evidence Gradient Plausibility Model exists for high dose effect Coherence Possible Experiment Not applicable Analogy No obvious case

From a public policy perspective, the plausibility of causality encourages consideration of the magnitude of the problem. Using a fairly crude and conservative approach, as many as 150-250 cases of childhood leukemia may be due to EMF exposure each year. This is a substantial number of cancers, and does not consider other cancers, other adverse outcomes or non-residential exposures. Some would argue that consideration of exposure reduction, if possible at limited cost and inconvenience, is warranted.

In light of these results, why do interpretations of causality and consequence differ among scientists, particularly among scientists from different disciplines? In part, I believe that there is substantial scientific uncertainty in the data. Each study has limitations, and subjective judgements are made about the size and importance of these limitations. In part, different scientists place different amounts of emphasis on different types of data. For example, in light of the uncertainties and inconsistencies, bench scientists are likely to give more emphasis to the in vivo and in vitro studies while epidemiologists give more emphasis to the epidemiologic

200 studies. Beyond that, the appropriateness, importance and interpretation of meta- analyses, risk analyses and other summaries are judged differently by different individuals. While most agree that the data are suggestive, gaining consensus beyond that is difficult as reflected in the many Blue Ribbon Panel reviews. As more studies are conducted and released, perhaps we will get greater clarity and consensus on this important issue.

References

1. Ahlbom A, Feychting M, Koskenvuo M, Olsen J, Pukkala E, Schulgen G, et al. Electromagnetic fields and childhood cancer. Lancet 1993; 342:1295-1296.

2. Blair A, Burg J, Foran J, Gibb H, Greenland S, Morris R, et al. Guidelines for application of meta-analysis in environmental epidemiology. Regulatory Toxicology and Pharmacology 1995; 22:189-197.

3. Cooper H. Statistically combining independent studies: A meta-analysis of sex differences in conformity. Journal of Personality and Social Psychology 1979; 37:131-146.

4. DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clinical Trials 1986; 7:177-188.

5. Foster K, Erdreich L, Moulder J. Weak electromagnetic fields and cancer in the context of risk assessment. IEEE Proceedings 1997; 85:733-746.

6. Greenland S. Evolution of Epidemiologic Ideas: Annotated Readings on Concepts and Methods. Chestnut Hill, MA: Epidemiologic Resources Inc., 1987.

7. Greenland S. Meta-analysis. In: Rothman KJ, Greenland S, eds. Modern Epidemiology. 2nd ed. Philadelphia: Lippincott-Raven, 1998:643-673.

8. Hill A. The environment and disease: Association or causation? Proceedings of the Royal Society of Medicine 1965; 58:295-300.

9. International Agency for Research on Cancer, World Health Organization. Mechanisms of Carcinogenesis in Risk Identification. Lyon, France: International Agency for Research on Cancer, 1992.

201 10. Kheifets L, Afifi A, Buffler P, Zhang Z, Matkin C. Occupational electric and magnetic field exposure and leukemia. Journal of Occupational and Environmental Medicine 1997; 39:(11)1074-1091.

11. Kheifets L, Afifi A, Buffler P, Zhang Z. Occupational electric and magnetic field exposure and brain cancer: A meta-analysis. Journal of Occupational and Environmental Medicine 1995; 37:(12)1327-1341.

12. Meinert R, Michaelis J. Meta-analyses of studies on the association between electromagnetic fields and childhood cancer. Radiation and Environmental Biophysics 1996; 35:15-18.

13. Miller M, Murphy J, M. T, Ruttenber A. Variation in cancer risk estimates for exposure to power frequency electromagnetic fields: A meta-analysis comparing EMF measurement methods. Risk Analysis 1995; 15:(2)281-287.

14. National Radiological Protection Board. Electromagnetic Fields and the Risk of Cancer. 1992; Chilton, Didcot, UK: National Radiological Protection Board. vol. 3: 138 p.

15. National Research Council. Risk Assessment in the Federal Government: Managing the Process. Washington, DC: National Academy of Sciences Press, 1983.

16. National Research Council. Possible Health Effects of Exposure to Residential Electric and Magnetic Fields. Washington, DC: National Academy Press, 1997.

17. NIEHS Working Group. Portier C, Wolfe M, eds. Assessment of Health Effects from Exposure to Power-Line Frequency Electric and Magnetic Fields. 1998; Research Triangle Park, NC: National Institute of Environmental Health Sciences, National Institutes of Health, US Department of Health and Human Services, Public Health Service. NIH Publication No. 98-3981: 508 p.

18. Olkin I. Invited Commentary: Re: "A critical look at some popular meta- analytic methods". American Journal of Epidemiology 1994; 140:(3)297-299.

202 19. Rosenthal R. The 'file drawer' problem and tolerance for null results. Psychological Bulletin 1979; 86:638-641.

20. United States Environmental Protection Agency. The risk assessment guidelines of 1986 (EPA/600/8-87/046). Federal Register 1986; 51:33992

21. United States Environmental Protection Agency. Proposed guidelines for cancer risk assessment. Federal Register 1996; 61:17960-18911.

22. Wartenberg D. Residential magnetic fields and childhood cancer: A meta- analysis. American Journal of Public Health 1998; in press:

23. Wartenberg D, Chess C. Risky business: the inexact art of hazard assessment. The Sciences 1992; 16-21.

24. Wartenberg D, Dietrich F, Goldberg R, Poole C, Savitz D. A Meta-Analysis of Studies of Childhood Cancer and Residential Exposure to Magnetic Fields. 1998; 53 pp. + tables p. A report to the National Institute of Environmental Health Sciences.

25. Wartenberg D, Greenberg M. EMF and The Media: Coverage by Three Major US Daily Newspapers. In: Matthes R, Bernhardt J, Repacholi M, eds. Risk Perception, Risk Communication and its Application to EMF Exposure. Oberschleissheim, Germany: International Commission on Non-Ionizing Radiation Protection, World Health Organization, 1998:215-228.

26. Washburn E, Orza M, Berlin J, Nicholson W, Todd A, Frumkin H, et al. Residential proximity to electricity transmission and distribution equipment and risk of childhood eukemia, childhood lymphoma, and childhood nervous system tumors: Systematic review, evaluation and meta-analysis. Cancer Causes and Control 1994; 5:299-309.

27. Zaffanella L. Survey of Residential Magnetic Field Sources. 1993; Palo Alto, CA: Electric Power Research Institute. TR-102759 (2 vols.):

28. Zaffanella L, Kalton G. Survey of personal magnetic field exposure; Phase I: Pilot study and design of Phase II. 1998; Oak Ridge, TN:

203 204 Assessment of health effects from exposure to power line frequency electric and magnetic fields

Christopher Portier National Institute of Environmental Health Sciences (NIEHS) Box 12233, Research Triangle Park, NC 27709, USA Tel: +1 919 541 4999, Fax: +1 919 541 1479 E-mail: [email protected]

and Mary Wolfe National Institute of Environmental Health Sciences Box 12233, MD F3-03, Research Triangle Park, NC, USA Tel: +1 919-541-7539, Fax: +1 919-541-0144 E-mail: [email protected]

Abstract

From June 16, 1998 to June 24, 1998, a group of thirty scientists both within and outside of electric and magnetic fields (EMF) research community met to examine the strength and robustness of the experimental data pertaining to EMF. The scientists came from a broad array of disciplines including epidemiology, biostatistics, toxicology, physics, cellular and molecular biology, pathology and mathematics. Their report covered the range of research available on EMF and drew a number of conclusions as part of a hazard identification for EMF health effects. This report is part of the evidence being used by the National Institute of Environmental Health Sciences to evaluate the potential for health risks from exposure to EMF. This brief report summarizes the conclusions from this Working Group.

INTRODUCTION

Research into the health effects of extremely low frequency (ELF) electric and magnetic fields (EMF) has progressed for greater than thirty years. The catalyst that sparked increased study in this area of research was the 1979 report by Wertheimer and Leeper that children living near power lines had an increased risk for developing cancer. Since that initial finding, there have been numerous epidemiological and laboratory studies aimed at clarification of this finding and

205 subsequent findings of similar concern. Despite this multitude of studies, there remains considerable debate over what, if any, health effects can really be attributed to ELF-EMF. In 1992, the U.S. Congress instructed the National Institute of Environmental Health Sciences (NIEHS) and the Department of Energy (DOE) to direct and manage a program of research and analysis aimed at ending this debate. This resulted in the formation of the EMF Research and Public Information Dissemination Program (EMF-RAPID). The program was funded jointly by Federal and matching private funds.

The EMF-RAPID Program has three basic components: 1) a research program focusing on health effects research primarily through mechanistic studies of ELF-EMF and engineering research targeting measurement, characterization and management of ELF-EMF; 2) information dissemination consisting of brochures, public outreach and a hotline for communicating with the public; and 3) a health assessment including an analysis of the research data aimed at summarising the strength of the evidence for evaluation of any hazard possibly arising from exposure to ELF-EMF. The NIEHS was directed to oversee the health effects research and evaluation and DOE has responsibility for the engineering research. The research program was a four year program which terminates in December, 1998. Much of the EMF-RAPID research is just now being published. The public outreach program developed a number of brochures which proved extremely useful as tools for informing the public about the nature of EMF and the types of research which address the potential for hazards.

The part of the EMF-RAPID Program aimed at the analysis and summarization of research findings began in 1996 with the development of a program consisting of a series of workshops addressing different areas of the research database. The first workshop was held in March, 1997 and reviewed the literature on in-vitro and mechanistic findings. Subsequent workshops focused on in-vivo and clinical research and on epidemiological findings. These workshops were aimed at including a broad spectrum of the research community in the evaluation of EMF health hazards, identifying key research findings and providing opinion on the quality of this research. A description of this series of workshops is given in Portier and Wolfe (1997b) and the risk communication strategy used in developing this series of workshops is given in Portier and Wolfe (1998a). Each workshop produced a report which is available from the NIEHS (Portier and Wolfe, 1997a, 1998b, 1998c).

206 Starting in late 1997, in preparation for an EMF-RAPID Working Group meeting, a group of select scientists were given the task of providing formal reviews of the literature in the areas of research identified from the three workshops. Eventually, twelve separate documents were prepared covering the areas of animal carcinogenicity, animal non-cancer findings, neuroendocrine effects, cell signaling effects, effects on DNA, effects on cellular replication, effects on cellular differentiation, biophysical theories, exposure, adult environmental epidemiology, childhood epidemiology and adult occupational epidemiology. These documents were distributed to thirty scientists chosen to form the NIEHS EMF Working Group. From 16-24 June, 1998, these scientists met in Brooklyn Park, Minnesota where they rewrote the draft documents into a single book (NIEHS Working Group Report, 1998). In addition to reviewing the literature, the Working Group also characterised the strength of the evidence in each category of research using the criteria put forward by the International Agency for Research on Cancer (IARC). These criteria are given in Appendix B of the Working Group Report.

WORKING GROUP FINDINGS

Levels of evidence

Evaluations of the various health end-points using the IARC criteria fall into four basic categories; sufficient evidence, limited evidence, inadequate evidence, and evidence suggesting the lack of an effect. Without repeating the definitions given in Appendix B of the Working Group Report, these categories can be viewed respectively as strong evidence generally including dose-response, weak evidence with an indication of dose-response, mixed evidence for negative findings from a limited number of studies and strong evidence of a no effect. Many of the effects covered in this report relate to ‘limited’ and ‘inadequate’ evidence. ‘Limited evidence’ is not an unusual finding for epidemiological data in the IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. This degree of evidence is generally provided by studies for which there is credible evidence of an association and for which a causal linkage cannot be established with a high degree of certainty. This does not mean the effect is weak (e.g. shallow dose-response), nor does it mean there is clearly an effect, although these issues enter into the evaluation. In most cases, this degree of evidence is associated with one or more of the following problems: questionable identification of the exposure factor(s) associated with the disease outcome (either a weak dose surrogate was used or individuals were misclassified as to their exposure category), bias may have played a small role in the finding, confounders were not ruled out to the satisfaction of the

207 original investigator and/or the Working Group, the observed effect was small, making clear detection of an effect difficult, and/or there is little information on dose–response in the available report. The careful reader is directed to the individual sections of the report for a clear description of each study. For experimental animal bioassays, ‘limited evidence’ for an effect is generally driven by clear findings in only a single study for a single end-point or minor problems with a set of data which otherwise would have been positive; confounding, bias, and exposure misclassification generally do not exist in laboratory studies.

‘Inadequate evidence’ can imply one of four possibilities: (1) there are insufficient data for making a judgment of any kind (e.g. poor study design, making interpretation impossible); (2) the data suggest a positive effect but, due to limitations in design or very weak findings, cannot be interpreted as suggesting a causal linkage; (3) the data suggest a negative effect but, due to limitations in design or very few findings, cannot be interpreted as suggesting no effect; and (4) the data are contradictory and no clear pattern is discernible. For case (1), given a solid hypothesis, it may be beneficial to continue to study an inadequate finding using a better design in the same experimental system. For case (2), if the effect seen is of public health consequence, it should be studied further but with a clear hypothesis and perhaps in conjunction with other studies such as those providing mechanistic interpretation. In case (3), unless there is a clear scientific reason for further study, again involving a defined hypothesis, there is little need to continue to study the observed effect. Finally, for case (4), the effect might be further studied if the scientific issues are compelling or if health concerns are raised, but it is unlikely that another study of similar design would be performed. Additional studies might not be needed. Again, a careful reader searching for scientific hypotheses for further study should read the more detailed descriptions of the findings presented in the Working Group Report.

Votes were taken for most of the conclusions given in the Working Group Report. The counts for each of these votes are reported in the text of the report. In most cases, the votes were nearly unanimous; however, some of the votes were very close. The range of votes is indicative of the degree of uncertainty among members concerning the strength of evidence for ELF-EMF associated health effects.

Carcinogenicity In Humans

The conclusions drawn by the Working Group on the evidence for carcinogenicity of ELF-EMF in humans are predominantly derived from

208 epidemiological studies which were designed to detect associations between health effects and exposure to EMF. This evidence is subdivided into adult occupational studies, adult residential studies and studies of childhood exposures. In each subdivision, the Working Group developed criteria mostly based upon experimental design and how fields were measured, for determining the inclusion or exclusion of studies in its evaluation. For each study included in the evaluation, the Working Group carefully assessed study quality, focusing on issues such as control selection, exposure assessment, confounding, and other possible sources of error which could lead to misinterpretation of the study findings. These issues are discussed in detail in the Working Group Report for each epidemiological study. The Working Group also evaluated six meta-analyses of childhood cancers, focusing on a recent analysis done by Wartenberg et al. (1998).

For adult occupational studies, the Working Group considered different cancers in the following categories: all cancers combined, leukaemia, brain cancer, breast cancer, lung cancer, central nervous system cancers in the offspring of parents occupationally exposed, and sporadic reports for other cancers. In addition, the Working Group reviewed meta-analyses of brain cancer and leukaemia. With the exception of chronic lymphocytic leukaemia, the Working Group concluded that there was inadequate evidence of carcinogenicity of ELF-EMF in adults. In most cases, this finding was driven by either a paucity of studies of adequate design to address the question of association or by mixed results with some studies showing a positive association and other studies showing no association with no predominant pattern.

Leukaemia was the first cancer associated with occupational exposure to EMF and has been investigated in over 70 studies, all of which were at least considered by the Working Group. Many were excluded because of limitations of the exposure metric or because better studies were available. In the one meta- analysis evaluated, a small but significantly increased relative risk for leukaemia and its main subtypes was found. In a careful evaluation of four studies on chronic lymphocytic leukaemia, three of the four studies showed increased relative risks (Feychting et al., 1997; Floderus et al. 1993; Theriault et al. 1994), although not all were statistically significant, and one showed no association (Savitz and Loomis, 1995). Taken together these studies suggest an association between EMF exposure and chronic lymphocytic leukaemia for which a causal interpretation is possible but not established; chance, bias or confounding could not be ruled out and may have led to the observed association. It should be noted that the Working Group based

209 this decision upon the bulk of this evidence and not on any one study since not all cohorts showed statistically significant results.

The evaluation of cancers associated with adult residential exposure focused mainly on leukaemia, breast cancer, and cancers of the central nervous system. As for occupational exposures, leukaemia was the initial cancer cited (Wertheimer and Leeper, 1987). Nine studies met the criteria for inclusion in this evaluation for an association between ELF-EMF and leukaemia. Breast cancers and nervous system cancers had fewer studies. In general, the Working Group felt there was inadequate evidence for an association between any of these cancers and adult residential exposure to ELF-EMF. These studies generally demonstrated no association or a weak association with no statistical significance which led to the Working Group’s opinion of inadequate evidence.

Wertheimer and Leeper (1979) initiated the hypothesis that EMF from electrical power lines and substations are associated with childhood cancer through their use of wire codes as a surrogate to characterise exposure. Their seminal paper led to a number of epidemiological studies investigating this hypothesis. These studies have examined either all cancers, leukemias (sometimes specific subtypes), lymphomas, or brain cancer. The Working Group focused on studies with measured or calculated magnetic fields giving less weight to studies using wire codes. With the exception of childhood leukaemia, the Working Group concluded there was inadequate evidence to support an association between ELF-EMF and childhood cancers. For childhood leukaemia, the Working Group concluded there is limited evidence for an association. The most compelling evidence supporting this finding came from four Nordic studies (Feychting and Ahlbom, 1993; Olsen et al. 1993; Verkasalo et al. 1993; Tynes and Haldorsen, 1997) in which calculated magnetic fields were used as the metric of exposure. Three of these studies found leukaemia risk increased with increasing calculated exposure while the smallest study (Tynes and Haldorsen, 1997) found no association. These findings were further supported by three studies with 24-hour magnetic field measurements (London et al., 1981; Linet et al., 1997; Michaelis et al., 1998) and by studies using wire-codes as exposure surrogates but not by studies using spot measurements. While all of these studies had limitations, the Working Group felt that chance was an unlikely explanation for the observed associations. It should also be noted that, for some studies, the conclusions of the Working Group did not coincide with the conclusions of the original authors.

Carcinogenicity in Animals

210 Animal carcinogenicity studies are routinely used to identify environmental agents which may increase cancer risks in humans. While animal studies are not plagued with many of the problems associated with epidemiology studies, they do present two additional problems: extrapolation across species and extrapolation from laboratory exposure patterns to environmental exposure patterns. In these studies, the potential for increased carcinogenic risks of EMF has usually been studied at intensities of exposure that are much higher than environmental exposures and have greater uniformity in frequency and intensity than would appear in environmental settings. These experimental conditions are chosen to maximize the ability for a researcher to detect an effect for a clearly defined exposure. However, because animal studies do not mimic the human situation, animal data are usually used to provide support for epidemiological data but seldom are used to refute that data. The IARC rules used by the Working Group (NIEHS Working Group, Appendix A, 1998) for ranking carcinogenic hazards approaches laboratory data in this fashion.

Several experimental designs and animal models have been used to study the potential carcinogenicity of ELF-EMF. In some cases, the choice of an animal model is based upon tradition whereas in other cases a model is chosen to address a specific mechanism. Lifetime studies of ELF-EMF as a complete carcinogen have been conducted in three separate studies (NTP, 1998b; Mandeville et al., 1997; Yasui et al., 1997). In two of these studies, the Working Group concluded there were no increases in cancer that could be attributed to EMF exposure. In the final study (NTP, 1998b), there was an isolated finding of an increase in thyroid C-cell adenomas and carcinomas in one sex-species group making this study equivocal according to the original report. One additional long-term study of 32 weeks duration also indicated no cancer effects from exposure to ELF-EMF (Marganato et al., 1995).

Mechanism-based models of carcinogenesis in mammary tissue, skin, leukaemia and liver were also evaluated. These studies showed inconsistent results across laboratories with a general trend toward no effect. Similar findings held for the few studies using transgenic animals. There was considerable debate over differences in the responses seen in a multistage model of mammary carcinogenesis done in a German Laboratory (Loscher et al. 1993, 1994; Mevissen et al. 1993, 1998) versus replicate studies done in a U.S. laboratory (NTP, 1998a). The U.S. studies were negative whereas the German studies showed some effects. A Swedish study (Ekstrom et al., 1997) using a similar protocol but slightly different exposure patterns was also negative. The Working Group concluded that this ensemble of

211 studies did not provide convincing evidence for a co-promoting effect of EMF on chemically-induced mammary carcinogenesis.

Following extensive debate and some degree of controversy, the Working Group voted 19 to 8 that there is inadequate evidence for carcinogenicity in animals (the 8 remaining votes were all for lack of carcinogenicity).

Non-Cancer Effects in Humans

Numerous epidemiological studies have looked at the possibility of health effects other than cancer arising from exposure to ELF-EMF. In general, the Working Group concluded the data were inadequate to support an association between ELF-EMF and any adverse health outcome other than cancer. These included adverse birth outcomes, reproductive effects, Alzheimer disease, lateral sclerosis, suicide, depression and cardiovascular disease. In most cases, these findings are severely limited by the experimental designs and the number of studies performed.

Laboratory studies of biological effects (non-adverse) were also evaluated. These were treated as mechanistic studies and evaluated using the categories weak, moderate and strong. It was concluded there was weak evidence that short-term human exposure to ELF-EMF causes changes in heart rate variability and sleep disturbance. There was no evidence that exposure to ELF-EMF had effects on any other biological endpoints studied clinically.

Of particular interest to scientists in this field is melatonin, a neuroendocrine hormone having multiple effects throughout the mammalian system. Several authors (e.g. Stevens et al., 1997) have proposed that modulation of melatonin production and release following exposure to EMF could explain some of the possible effects seen in the literature. Clinical and epidemiological studies were available for evaluating this hypothesis. The Working Group concluded there was weak evidence to support an effect of ELF-EMF on circulating melatonin levels but drew no conclusions concerning the linkage between altered melatonin and health risks.

Non-Cancer Effects in Animals

Non-cancer effects studied in laboratory animals cover a broad range of experimental animals, study designs, exposures and biological effects. Most of these endpoints have been evaluated in multiple studies with more than one species. The

212 review of these studies covered over 60 pages in the Working Group report with detailed evaluation of over 100 studies.

An extensive range of assay systems has been developed to evaluate the effects of environmental agents on the immune system. These assay systems are targeted at different aspects of the immune system; most notably acquired (humoral or cellular) and innate immunity. Animal models studied include baboons, rats and mice. Two research groups (House et al., 1996; Tremblay et al., 1996) reported significant effects on NK cell acitivity, but in different directions and with mixed results when repeated in one laboratory (House et al., 1996). The studies had varied quality and suffered from experimental difficulties. The Working Group concluded there is no evidence for effects of ELF-EMF on the immune system in experimental animals.

Hematological effects of ELF-EMF included measures of red blood cells, hemoglobin concentration, red cell volume, corpuscular volume (of various sorts), hematocrit and cell number (various cell types). In addition, bone marrow cellularity and bone marrow smears were also evaluated. The Working Group concluded there was no evidence for effects of ELF-EMF on hematological parameters in rodents.

There is little doubt that animals can detect electric fields with detection thresholds in the range of 3-10 kV/m. This is a well-established phenomenon supported by numerous studies and can be shown to alter behavior through studies of avoidance, aversion, performance and learning. The Working Group found no evidence suggesting the same was true for magnetic fields at environmentally - relevant exposures. ELF-EMF effects on neurophysiology and electrophysiology were not as clear. The Working Group concluded there is weak evidence for effects of electric and magnetic fields on neurobehavior, neuropharmacology, neurophysiology, and neurochemistry in experimental animals.

Reproductive and developmental effects of ELF-EMF have been studied in birds, mice, rats and hamsters. A few laboratories have reported alterations in the development of chick embryos but the results of teratogenic and reproductive effects in mammalian systems have generally been negative. No studies were available on other developmental endpoints. The Working Group concluded there is no evidence for reproductive and developmental effects of exposure to ELF-EMF in experimental animals.

213 Several studies were specifically designed to evaluate the melatonin hypothesis; most of these studies were done in either rats or hamsters. Study designs included both short-term and long-term exposures. The evidence for a quantitative and consistent effect is lacking but generally supported a slight depression in melatonin in rats. High exposure and field orientation appear to play an important rule in modulating this response. It is unclear what biological significance may be attributed to this reduction. The Working Group concluded there was weak evidence that exposure to EMF alters the levels of melatonin in rodents, but there was no evidence supporting such an effect in sheep or baboons.

For several years, EMF has been used as a non-invasive therapy for the promotion of bone healing. In this clinical use, it is believed that the induced electric field is the agent stimulating healing. Numerous studies in humans and experimental animals were reviewed by the Working Group. Additional research has been done investigating the possibility that EMF can also heal damage to soft connective tissue (skin and ligaments) and nerve tissue. The Working Group concluded there is strong evidence that exposure to electric and magnetic fields affects bone repair and adaptation. The Working Group could not reach a conclusion for the other two tissues.

In-Vitro and Mechanistic Studies

The number of cellular components, processes and systems which can possibly be affected by ELF-EMF is very large. Most of the studies in this area have focused on the cell membrane, gene expression and signal-transduction pathways. Initial research also evaluated genotoxicity of ELF-EMF but was mostly negative; recent research has renewed efforts to study possible roles of EMF in causing DNA damage. Most of the experimental evidence in this area arises from studies using cultured cells exposed in-vitro. The Working Group evaluated a broad range of studies in this category of research, focusing on studies designed with a clear mechanistic hypothesis. In evaluating these studies, the Working Group considered three critical factors: independent validation of a finding, demonstration of a physical mechanism supporting the observed effect, and an understanding of a linkage between the observed effect and an adverse health outcome.

The Working Group highlighted a series of recent studies (1996-98) demonstrating effects of EMF on gene mutations. They noted that studies of ELF- EMF exposures below 0.1 mT have consistently shown no effect on mutation rates; however, exposures of 0.2-400 mT produced significant enhancement of mutation

214 after x-ray and gamma-ray initiation (Walleczek et al., 1998; Miyakashi et al., 1996, 1998). In addition, exposure at 400 mT induced mutations in the absence of ionising radiation (Miyakashi et al., 1996). Some studies were found to produce strong effects in other endpoints commonly associated with cancer (e.g. cell proliferation, disruption of signal transduction pathways and inhibition of differentiation). These studies demonstrated effects at exposures associated with a plausible mechanism (magneto-chemical transduction; exposure > 0.1mT).

One set of studies received considerable attention at both the Working Group Meeting and at subsequent meetings: the studies looking into a possible reversal of the oncostatic effects of melatonin and tamoxifen by low intensisty (1.2 µT) ELF-EMF exposures (Liburdy et al., 1993; Harland and Liburdy, 1997; Blackman et al., 1998, Liburdy and Levine, 1998). While these studies are intriquing, the Working Group members were concerned about possible design flaws in these studies and the very small effect observed. Additional studies should soon be reported clarifying this issue.

The Working Group concluded that a limited number of well-performed studies provide moderate evidence for mechanistically-plausible effects of EMF at exposures exceeding 0.1 mT. There was weak evidence for effects below this exposure level.

The Working Group also summarized the biophysical mechanisms which have been hypothesised to support or refute the possibility of ELF-EMF biological effects. No conclusions were drawn from this review; the pros and cons for each theory were described. In summary it was noted that multiple theories may be applicable at different field intensities in different species; thus several models may be appropriate depending upon the case being studied.

DISCUSSION

The Working Group was asked to provide an overall evaluation for the carcinogenicity of ELF-EMF in humans. The best way to summarise their overall finding is to simply quote it:

“A majority of the Working Group concluded that classification of ELF- EMF as possibly carcinogenic (Group 2B) is a conservative, public-health decision based on limited evidence of an increased risk for childhood leukemias with residential exposure and an increased occurrence of CLL associated with

215 occupational exposure. For these particular cancers, the results of in-vivo, in-vitro, and mechanistic studies do not confirm or refute the findings of the epidemiological studies. The overall body of evidence has, however, laid a foundation for furthering our understanding of the biological effects, mechanisms, and exposure circumstances that may be related to the possible carcinogenicity and other adverse human health effects of exposure to ELF-EMF.”

This is not the end of the evaluation for ELF-EMF but instead the beginning. The Working Group Report is a recommendation to the NIEHS on the strength of the evidence for health effects from exposure to ELF-EMF. The NIEHS is currently seeking comment on this report as it prepares to develop and write a separate report (the NIEHS Director’s Report) mandated under the 1992 Energy Policy Act. To this end, the NIEHS has distributed over 1000 copies of the Working Group Report to scientists, regulators, industry and the public. In addition, the NIEHS has hosted a series of public meetings to receive comments on this document. At each meeting, NIEHS officials and other scientists involved with preparation of the NIEHS Director’s report were present to receive comments on the report. Attendees had the opportunity to voice their opinions and comments on the report to the panel orally and/or in writing at these meetings. In addition, the public meetings were preceded by poster sessions describing the NIEHS and its mission, the EMF-RAPID Program and the major findings from the Working Group Report.

All aspects of the Working Group report are important to the NIEHS and our considerations of the potential health effects from EMF exposure, but some may have more impact than others. One aspect of critical importance is the conclusion by a majority (19) of the Working Group members that EMF should be considered a “possible human carcinogen”. As mentioned earlier, the vote for possible carcinogenicity was based largely on epidemiological evidence in the face of animal and other laboratory studies that the panel agreed did not support or refute the population studies. No members of the Working Group concluded that EMF could be classified as a “known human carcinogen” (IARC class 1) or as a “probable human carcinogen” (IARC class 2A). The largest minority (8 members) concluded that EMF was “not classifiable as to its carcinogenicity” (IARC class 3) and one member concluded EMF was “probably not a human carcinogen” (IARC class 4). These conclusions about the possible carcinogenicity of EMF is an area where we are encouraging opinion.

There has been considerable debate about how the conclusions in this report compare with a recent National Academy of Sciences (NAS) Report (NAS, 1997).

216 The NAS Report stated that “no conclusive and consistent evidence shows that exposures to residential electric and magnetic fields produce cancer” (page 2, NAS Report). As noted above,the Working Group also did not conclude that EMF is a “known human carcinogen” (IARC class 1) or a “probable human carcinogen” (IARC class 2A) which is consistent with the NAS Report. However, the NAS and Working Group reports do differ in their assessment of the level of evidence from epidemiological studies linking EMF exposure and childhood leukaemia. The NAS concluded “Average magnetic fields measured in the homes of children have not been found to be associated with an excess in childhood leukaemia” (page 117, NAS Report). In contrast, the Working Group concluded the data “support an association between exposure to calculated magnetic fields and the incidence of childhood leukaemia” (page 186, Working Group Report) and give “some support for a possible association between exposure based on 24-h measured magnetic fields and the incidence of childhood leukaemia” (page 187, Working Group Report). This assessment of the epidemiological data led a majority (20) of the Working Group to conclude that there is “limited evidence” that EMF is carcinogenic to children (page 189, Working Group Report). The remaining members who voted (6) concluded the evidence was “inadequate”. This vote of “limited evidence” in children drove the Working Group’s classification of EMF as a “possible human carcinogen” since, according to IARC criteria (page 499, Working Group Report), one way in which an agent falls into the 2B category is if “there is limited evidence of carcinogenicity in humans and less than sufficient evidence of carcinogenicity in experimental animals.”.

The remaining steps in the process used by the NIEHS will include comment from the Federal regulatory community and final preparation of the NIEHS report for Congress. It is clear that some evaluation of the possible risks for childhood cancer from exposure to ELF-EMF will need to be given under the assumption that the association seen in the epidemiology studies is causal. This is not to inply a zero risk is unlikely, but simply a means to gauge the magnitude of any potential health risks which might arise.

REFERENCES

Ekström, T., Mild, K.H. & Homberg, B. (1998). Mammary tumours in sprague- dawley rats after initiation with dmba followed by exposure to 50 Hz electromagnetic fields in a promotional scheme. Cancer Letters, 123, 107-111.

217 Feychting, M. & Ahlbom, A. (1993). Magnetic fields and cancer in children residing near Swedish high-voltage power lines. American Journal of Epidemiology, 138, 467-481.

Feychting, M., Forssen, U. & Floderus, B. (1997). Occupational and residential magnetic field exposure and leukaemia and central nervous system tumours. Epidemiology, 8, 384-389.

Floderus, B., Persson, T., Stenlund, C., Wennberg, A., Ost, A. & Knave, B. (1993). Occupational exposure to electromagnetic fields in relation to leukaemia and brain tumours: a case-control study in Sweden. Cancer Causes Control, 4, 465- 476.

Harland, J.D. & Liburdy, R.P. (1997). Environmental magnetic fields inhibit the antiproliferative action of tamoxifen and melatonin in a human breast cancer cell line. Bioelectromagnetics, 18, 555-562.

House, R.V., Ratajczak, H.V., Gauger, J.R., Johnson, T.R., Thomas, P.T. & McCormick, D.L. (1996). Immune function and host defense in rodents exposed to 60 Hz magnetic fields. Fundamental Applied Toxicology, 34, 228-239.

Liburdy, R.P., Callahan, D.E., Harland, J., Dunham, E., Sloma, T.R. & Yaswen, P. (1993). Experimental evidence for 60 Hz magnetic fields operating through the signal transduction cascade. effects on calcium influx and c-myc messenger- RNA induction. FEBS Letters, 334, 301-308.

Liburdy, R.P. & Levine, G.A. (1998). Magnetic fields and formation of organized structures in normal human mammary cells. In BEMS Annual Meeting.

Linet, M.S., Hatch, E.E., Kleinerman, R.A., Robison, L.L., Kaune, W.T., Friedman, D.R., Severson, R.K., Haines, C.M., Hartsock, C.T., Niwa, S., Wacholder, S. & Tarone, R.E. (1997). Residential exposure to magnetic fields and acute lymphoblastic leukaemia in children. New England Journal of Medicine, 337, 1- 7.

London, S.J., Thomas, D.C., Bowman, J.D., Sobel, E., Cheng, T.-C. & Peters, J.M. (1991). Exposure to residential electric and magnetic fields and risk of childhood leukaemia. American Journal of Epidemiology, 134, 923-937.

218 Margonato, V., Nicolini, P., Conti, R., Zecca, L., Veicsteinas, A. & Cerretelli, P. (1995). Biologic effects of prolonged exposure to ELF electromagnetic fields in rats: II. 50 Hz magnetic fields. Bioelectromagnetics, 16, 343-355.

Michaelis, J., Schuz, H., Meiner, R., Zemann, E., Grigat, J.-P., Kaatsch, P., Kaletsch, U., Miesner, A., Brinkmann, K., Kalkner, W. & Karner, H. (1998). Combined Risk Estimates for Two German Population-Based Case-Control Studies on Residential Magnetic Fields and Childhood Acute Leukaemia. Epidemiology, 9, 92 - 94.

Miyakashi, J., Koji, Y., Wakasa, T. & Takebe, H. (1998). Long-term exposures to magnetic field (5mT at 60 Hz) do not increase mutations, but slightly enhance X-ray-induced mutations. Radiation Research, In press.

Miyakoshi, J., Ohtsu, S., Shibata, T. & Takebe, H. (1996). Exposure to magnetic field (5 mT at 60 Hz) does not affect cell growth and c-myc gene expression. Journal of Radiation Research (CHIBA), 37, 185-191.

National Academy of Sciences (NAS) (1997) Possible Health Effects of Exposure to Residential Electric and Magnetic Fields, National Academy Press, Washington, DC.

NIEHS Working Group (1998) Assessment of Health Effects from Exposure to Power-Line Frequency Electric and Magnetic Fields, Portier, C. and Wolfe, M. (Eds). Research Triangle Park, NC: National Institue of Environmental Health Sciences: Durham, NC.

Olsen, J.H., Nielsen, A. & Schulgen, G. (1993). Residence near high voltage facilities and risk of cancer in children. Britrish Medical Journal, 307, 891-895.

Portier, C., & Wolfe, M., eds. (1997a). EMF Science Review Symposium Breakout Group Reports for Theoretical Mechanisms and In Vitro Research Findings. Research Triangle Park, NC: National Institue of Environmental Health Sciences: Durham, NC.

Portier, C. and Wolfe, M. (1997b) Linking science to decisions: a strategy for electric and magnetic fields. In Matthes, R., Bernhardt, J. and Repacholi, M. (ed.) Proceedings of the ICNRP/WHO symposium on biological effects of ELF

219 electric and magnetic fields and related health risks, ICNRP Press, Germany. 211-218.

Portier, C. and Wolfe, M. (1998a) Risk Communication: Focus in the NIEHS RAPID review of EMF hazards. In Matthes, R. (ed.) Proceedings of the ICNRP/WHO symposium on Risk Perception, Risk Communication and its Application to EMF, ICNRP Press, Germany. 295-301.

Portier, C., & Wolfe, M., eds. (1998b). EMF Science Review Symposium Breakout Group Reports for Epidemiolgical Research Findings. Research Triangle Park, NC: National Institue of Environmental Health Sciences: San Antonio, TX.

Portier, C., & Wolfe, M., eds. (1998c). EMF Science Review Symposium Breakout Group Reports for Clinical and In Vivo Laboratory Findings. Research Triangle Park, NC: National Institue of Environmental Health Sciences: Pheonix, AZ.

Savitz, D.A. & Loomis, D.P. (1995). Magnetic field exposure in relation to leukaemia and brain cancer mortality among electric utility workers. American Journal of Epidemiology, 141, 123-134.

Stevens, R.G., Wilson, B.W. & Anderson, L.E. (1997). The Melatonin Hypothesis: Breast Cancer and Use of Electric Power pp. 1-760. Battelle Press: Columbus.

Tremblay, L., Houde, M., Mercier, G., Gagnon, J. & Mandeville, R. (1996). Differential modulation of natural and adaptive immunity in fischer rats exposed for 6 weeks to 60 Hz linear sinusoidal continuous-wave magnetic fields. Bioelectromagnetics, 17, 373-383.

Tynes, T. & Haldorsen, T. (1997). Electromagnetic fields and cancer in children residing near norwegian high-voltage power lines. American Journal of Epidemiology, 145, 219-226.

Verkasalo, P.K., Pukkala, E., Hongisto, M.Y., Valjus, J.E., Jarvinen, P.J., Heikkila, K.V. & Koskenvuo, M. (1993). Risk of cancer in Finnish children living close to power lines. British Medical Journal, 307, 895-898.

Walleczek, J., Shiu, E. & Hahn, G.M. (1998). Increase in radiation-induced HPRT gene mutation frequency from nonthermal exposure to non-ionising 60 Hz electromagnetic fields. Radiation Research, In press, 1-30.

220 Wartenberg, D., Dietrich, F., Goldberg, R., Poole, C. & Savitz, D. (1998). A meta- analysis of studies of childhood cancer and residential exposure to magnetic fields. Report for the National Institute of Environmental Health Sciences: Research Triangle Park, NC.

Wertheimer, N. & Leeper, E. (1979). Electrical wiring configurations and childhood cancer. American Journal of Epidemiology, 109, 273-284.

Wertheimer, N. & Leeper, E. (1987). Magnetic field exposure related to cancer subtypes. Annals of the New York Academy of Science, 502, 43-54.

221 222 Public Participation: Guiding Practice With Research

Caron Chess Center for Environmental Communication, Cook College, Rutgers University 31 Pine Street, New Brunswick, NJ 08903 Tel: (732)932-8795, Fax: (732)932-7815 E-mail: [email protected]

INTRODUCTION

One often-cited article calls public participation a “practice in search of a theory” (Wengart, 1976). Fortunately, theoretical frames for public participation are now discussed seriously although there is by no means consensus about a theory of public participation (e.g., Fiorino, 1990; Laird, 1993; Renn et al, 1995).

On the other hand, empirical research on public participation is sorely lacking. Agency practitioners who conduct public participation use guidance documents that are based largely on wisdom from accumulated experience (e.g., DOD and EPA, 1994). While this experience is invaluable, it is not a substitute for empirical research. But, practitioners have relatively little choice. There has been little systematic research to guide public participation practice (NRC, 1996), and much of the seminal research in the U.S. dates before the mid-1980s. In short, public participation is also a practice in search of research.

The usefulness of the relatively small research literature is also further hampered because studies often use vastly different criteria to explore public participation efforts. Some studies define public participation in terms of public participation process, that is “how” a public participation program is conducted (Tuler and Webler, 1995). Others look at the substantive outcome of a public participation effort (e.g., progress in remediation, control of non-point source pollution, etc.) and others explore both (e.g., DOE, 1996).

In addition, goals for evaluating public participation reflect conflicting values about and expectations for public participation. For example, three handbooks on public participation agree little on criteria for evaluating models of public participation (Webler, 1997). The peer-reviewed literature is equally heterogeneous and mentions, among other goals, acceptance of agency plans,

223 consensus, improvement of agency plans, social learning, public education about environmental problems, and agency awareness of citizen concerns.

In this paper I draw on research about public participation in North America to discuss implications for practice. I explore the assumptions currently made about public participation and the extent to which those assumptions are substantiated with the empirical research. (For a more comprehensive review, see Chess and Purcell, 1998, which raises similar issues.)

1. Timing of participation

The public participation rule of thumb suggests that public participation should be conducted early in the decision-making process (e.g., Kasperson, 1986). Although empirical research is far from definitive, investment in these preliminary stages of public participation appears prudent. Studies faulted the timing of the participatory effort when participants were asked to react to agency plans--the proverbial “decide, announce, and defend” approach--rather than join in earlier discussions of alternatives (e.g., Gariepy, 1991; Mazmanian and Nienaber, 1979). Insufficient or inappropriate outreach was cited as a problem in some public meetings (Sinclair, 1977). Two studies of workshops also found that workshop participants wanted a greater role in planning participation (Brandt, 1994; Young et al., 1993). Similarly, the success of CACs may depend on agencies’ willingness to involve participants in all phases of the planning process (Lynn and Busenburg, 1995).

2. Relationship between the process and the outcome of public participation

“People are more willing to accept the results of processes they perceive as fair, balanced and reasonable and that allow them an adequate opportunity to have a fair say,” according to a report from the National Research Council (1996). The NRC report backs this assertion with citations of case studies that tie involvement of stakeholders in decision making with reducing conflict about the siting of hazardous waste sites (e.g., Kraft, 1988; Heiman, 1990).

Because siting can be seen as an extreme test of public participation due to the extent and nature of the conflict, the NRC’s evidence of strong links between process and outcome are arguably particularly significant. If improved process can engage participants in dialogue about hosting a locally unwanted land use, then

224 improved process might have an even greater impact on less controversial environmental issues.

In fact, many studies show a link between participants’ satisfaction between process and outcome. On the other hand, four carefully conducted studies suggest that although stakeholders may respond favorably to a participation process, they may not respond as favorably to the outcomes. For example, participants rated an Army Corps workshop that explored wetlands permitting satisfactory on process but unsatisfactory on outcomes, leading the researcher to suggest that “participants’ support for the process does not lead to immediate support for outcomes” (Rosener, 1981:594). Another early, but much cited, study of participatory efforts of the Corps of Engineers described a case where the final meeting found both Corps and citizen participation self-congratulatory (and the process ratings more than satisfactory), yet participants rated the outcome as less than satisfactory (Mazmanian and Nienaber, 1979).

A more recent study surveyed advisory boards at major DOE sites on both process and outcome and found more overall satisfaction with the process of citizen involvement than with the outcome (U.S. DOE, 1996). For example, based on combined responses to 13 items, 71% of respondents agree or strongly agree that SSABs facilitate interaction and exchange of viewpoints on SSAB site issues. Yet based on combined responses to six items, a much less substantial majority (53%) of respondents agree or strongly agree that the SSABs lead to more acceptable agency actions.

Surprisingly a study of two military advisory boards found that members of one were positive about the outcome despite discontent with elements of the CAC process, including poor facilitation (Beltsen, 1996).

Based on this evidence, the link between process and outcome should not be assumed. On the other hand, without a positive participation process, dissatisfactions with outcome might have been even greater. Further research is critically needed to explore the association between process and outcome.

3. Selection of participatory form

There have been attempts to characterise forms of participation (e.g., public meetings, citizen advisory councils, citizen juries, surveys, etc.) in a way that makes explicit when they should be used. However, the NRC (1996) concluded that “there

225 is no rigorously or generally accepted classification scheme” to predict which participatory form will work in any given situation. In other word, selecting the “correct” form of participation may not be the factor that most shapes the outcome of a given participatory effort. Briefly, I discuss this concept in terms of two forms of participation: public meetings and citizen advisory committees.

Public meetings

For years practitioners have been advised to reduce the use of public meetings (e.g., English et al., 1993). The two most prominent normative theories of public participation suggest that public meetings are not particularly useful because they do not promote competence and fairness (Renn et al., 1995) or promote face-to- face discussions over time (Fiorino, 1990). Agency staffers can be equally derogatory about public meetings, as exemplified by one agency staffer who defined his goal for an upcoming public meeting as “to survive” (Chess et al, 1989).

Public meetings have also been disparaged for other reasons, including the lack of representativeness of opinions or the demographics of those at meetings (e.g., English, 1993). However, some empirical research undercuts this generally- accepted public participation truism. For example, opinions based on responses to randomized surveys and comments at public meetings were found to be comparable (O’Riordan, 1976; Gundry and Heberlein, 1984). Analysis of meeting transcripts about transportation plans found considerable diversity in reasons for and intensity of concern, suggesting that merely counting opponents and proponents oversimplifies the complexity of issues and the nature of representation (Kihl, 1985). Interestingly, surveys of four public participation efforts held by the Corps of Engineers found that people preferred public meetings over workshops and seminars (Mazmanian and Nienabur, 1979).

Another charge leveled at public meetings is that they legitimate agency decisions that have already been made (Fiorino, 1990; Checkoway). However, some studies suggest that participation can have important impacts on plans (e.g. Rosener, 1982; Elder, 1982). Most notably, public meetings have served to block agency decisions due to overwhelming opposition (e.g., Mazmanian and Nienaber, 1979; Gariepy, 1991).

While public meetings have come under increasing attack, citizen advisory councils (CACs) are ascendant. Chemical manufacturers, ranging from small companies to multinational corporations have created hundreds of them over the

226 past 10 years. In addition, the U.S. Departments of Defense and of Energy have both mandated development of site-specific advisory boards to explore remediation of contamination at the agencies’ facilities.

Yet CACs have also had their share of critics who argue that advisory committees are often unrepresentative, require large time commitments, and may not have legitimacy in the eyes of other citizens (English et al, 1993). A review of 14 studies of CACs points to examples "where broadly based CACs, with well-defined charges, adequate resources, and neutrally facilitated processes had significant policy impacts" (Lynn and Busenberg, 1995:159). However, the authors caution that the CACs’ influence was "highly contingent" on the agencies’ intentions. For example, when the agency’s goal was merely gaining support for the agency or mollifying critics, CACs were less likely to have an impact. If agencies truly wanted input, then CACs were more effective.

The results of these studies are sufficient to raise questions about how much the planning of public participation should focus on selection of the “correct” form of public participation. Process or outcome results may be due as much to how an agency uses these forms as the form itself. For example, public meetings take less agency time than CACs that usually meet on a routine basis. Agencies that care little for public input are likely to take the fastest approach - often public meetings - and then, not surprisingly, fail to find the forum useful. In addition, public meetings are characterised as promoting posturing and increasing conflict. However, from the research studies it is unclear whether the timing of public meetings, often after the “decide and announce” phases, is the problem or the form itself is inherently flawed.

Conversely, CACs are seen, in general, as allowing for more interaction, information exchange, and growth of mutual understanding. This may be so, in part, because agencies convened the CACs before decisions were made. In short, forms of public participation are tools; their success may depend as much as how they are used as whether they are the right tool for the job. While no doubt different forms of participation may have different strengths and limitations, these characteristics may not determine either process or outcome success - the history of the issue, level of conflict, scientific data, and existing power dynamics may influence outcome as much as the method” (NRC, 1996).

4. Modification of participatory forms

227 Instead of characterizing a participatory process as appropriate for some situations and not others, modifying the form may provide agencies with more options. In the case of public meetings, it is possible to hold them earlier in the decision making process or to develop meeting agendas that include presentation or generation of alternative proposals, as requested by participants in one Corps effort (Mazmanian and Nienaber, 1979). Group discussions can serve as a part of, or supplements to, public meetings (e.g. Sinclair, 1977). Some agencies have also experimented with meetings in which agency personnel serve as resources on specific topics at dialogue stations around the room, allowing participants to visit the station that most interests them (Hance and Chess, 1996). Others have used neutral facilitators (e.g., Young et al, 1993) or have co-sponsored meetings with non- government organizations to improve on the meeting process.

New media are also facilitating different approaches to meetings. For example, the Alaskan Department of Transportation created a dynamic public meeting soliciting input via telephone and consensor, a hand held electronic device hooked to a computer that tallied responses to questions in 3-4 seconds. While the 4,000 participants were not demographically representative of the seven urban areas in which this experiment took place, more than 50% of the 95,000 people who watched the program had observed a public meeting for the first time. (Slaton, 1992). Particularly promising are innovative participatory programs that are designed with multiple participatory forms, such as public meetings and citizen advisory councils, each involving different audiences.

5. Agency action

Some empirical research suggests that agency actions (or inactions) such as delayed requests for stakeholder input, failure to publicize forums, lack of technical support, and lack of commitment are associated with process and outcome limitations. Agencies also contributed to success, according to researchers, by providing technical assistance (e.g., Elder, 1982), providing experienced facilitation (Beltsen, 1995), and committing to the process (e.g., Rosener, 1981; Lynn and Busenberg, 1995), among others.

The question of agency commitment has been a recurring theme in discussions of public participation. For example, practitioners and academics at a national symposium were so concerned about agency resistance that they called for the agencies themselves to be a major topic of research (Chess et al, 1995). In addition, investigations of interactive forms of public relations, in which both the

228 organization and outside stakeholders change in response to each other, has found that the structure and staffing of organizations is associated with its responsiveness (Dozier and Grunig, 1992). Also, study of corporate efforts to involve community stakeholders suggests that organizational factors both facilitated and impeded these efforts. In particular, the organizational strength of the functional links between what an organization says to stakeholders and what it actually does may be particularly important (e.g., Chess, 1997). Evaluators of policy also underscore the importance of exploring institutional variables, “... what is evaluated is political architecture, architecture that influences outcomes, structures, processes, and constructs...” (Bartlett, 1994).

6. Evaluating public participation

While many articles call for more research on environmental public participation, more attention needs to be paid to how to do so.

Form of evaluation

Most evaluations of public participation are retrospective, which can help agencies consider whether to expand, modify or terminate programs. Retrospective evaluations also provide invaluable contributions to knowledge about public participation. However, such hindsight does not provide agency managers with feedback to make midcourse corrections in their participatory programs.

On the other hand, formative evaluations, which take place during program design (e.g., in the form of needs assessments) or during program development (e.g., routine evaluations of meetings), are aimed at improving efforts in progress. Formative evaluations can be seen as analogous to medical testing that takes place before treatment and periodically after initial treatment and diagnosis. Just as a physician can decide to alter treatment based on test results, formative evaluation provides managers with feedback during program development and implementation (Posavac, 1991).

To improve programs as they evolve, formative evaluation can consider complex issues such as how well agencies (or units in agencies) are cooperating, what kind of data needs to be recorded to track a program, where resources are flowing, and how implementation differs among sites (Chelimsky, 1997), as well as more obvious concerns such as the relationships among stakeholders, perceptions of agency communication, the effectiveness of meetings, etc. Formative evaluations

229 which involve participants in design and/or implementation can also help citizen initiatives stay on track (e.g., through workbooks that help groups determine their goals and if they are meeting them) (Linney and Wandersman, 1996).

Participatory evaluation

Currently, one of the most contentious arguments in the evaluation field is about the extent that program personnel and stakeholders should be involved in evaluation. Evaluators' opinions range from maintaining objectivity through distancing (e.g., minimising interviewing if it is possible to use extant data) (Scriven, 1997) to empowering stakeholders to design and implement their own evaluations (Fetterman, 1996).

In contrast to the argument for as much separation as possible between evaluator and program personnel, a participatory perspective views an evaluator as an educator, who encourages learning, rather than a referee who determines right from wrong. Participatory evaluation exemplifies this process by involving agency staff and other stakeholders in the design and/or implementation of evaluation. According to advocates of this type of evaluation, this process increases the likelihood that the evaluation will be viewed as credible and useful because the diverse needs of participants are more likely to be fulfilled (e.g. Greene, 1987; Guba and Lincoln, 1989).

DOE's assessment of its Site Specific Advisory Boards is an example of participatory evaluation. A work group of representatives of headquarters' managers, site staff and stakeholders, advised by consultants and academics, developed the evaluation design; outside consultants administered a survey and conducted the data analysis. The rationale for involving these participants included the desire to develop an evaluation design that would yield information helpful to "program implementors and decision makers," ground the evaluation in "mutually understood and agreed upon goals," take into account different participants' ideas of the actions that would lead to achieving these goals, and respond to local needs and processes. (DOE, 1996: 4-5).

Agency practitioners and academic researchers, who gathered at a 1994 symposium about participatory forms of risk communication, focused on ways to encourage agency evaluation (Chess et al, 1995). Discussions of organizational impediments to evaluation took precedence over methodological ones. Both practitioners and researchers expressed concern that the lack of documentation

230 makes it difficult to demonstrate to managers the usefulness of communication. However, the supporting data are not collected in part, due to limited management support, according to symposium participants. DOE’s effort to evaluate its site- specific advisory boards is an innovation other agencies are considering. If they do emulate DOE, public participation may become, over time, a practice supported by research.

REFERENCES

Beltsen, L. (1996). Assessment of local stakeholder involvement. Western Governors' Association

Checkoway, B. (1981). The politics of public hearings. The Journal of Applied Behavioral Science, 17(4):566-581.

Chelimsky, E. (1997). The Coming Transformations in Evaluation. In E. Chelimsky & W. Shadish (Eds.), Evaluation for the 21st Century (pp. 1-26). Thousand Oaks, California: Sage Publications.

Chess, C. (1997). Risk communication as an organizational response: Cases of Three Chemical Manufacturers [diss], Syracuse: State University of New York Environmental Science and Forestry.

Chess, C., Hance, B.J., & Sandman, Peter. (1989). Planning dialogue with communities: A risk communication workbook.

Chess, C., Salomone, K., Hance, B. J., & Saville, A. (1995). Results of a national symposium on risk communication: Next steps for government agencies. Risk Analysis, 15(2), 115-125

Dozier, David M. and Grunig, L. (1992) The organization of the public relations In J. E. Grunig, D. M. Dozier, W. P. Ehling, L. A. Grunig, F. C. Repper & J. White (Eds.), Excellence in public relations and communication management. Hillsdale: Lawrence Erlbaum

Elder, P. (1982). Project approval, environmental assessment and public participation. The Environmentalist, 2(1):55-71.

231 Federal facilities environmental restoration dialogue committee. (1996). Final report of the Federal facilities Environmental Restoration Advisory Committee. Federal Facilities Restoration and Reuse Office, USEPA.

Fetterman, D. M. (1996). Empowerment evaluation: An introduction to theory and practice. In D. M. Fetterman, S. Kafterian, J. Kafterian & A. Wandersman (Eds.), Empowerment evaluation: Knowledge and tools for self-assessment and responsibility (pp. 3-45). Thousand Oaks: Sage.

Fiorino, D. J. (1990). Citizen participation and environmental risk: A survey of institutional mechanisms. Science, Technology, and Human Values, 15, 226-243

Gariepy, M. (1991). Toward a dual-influence system: assessing the effects of public participation in environmental impact assessment for hydro-Quebec projects. Environmental Impact Assessment Review, 11:353-374.

Greene, J. (1987). Stakeholder participation in evaluation design: Is it worth the effort? Evaluation and Program Planning, 10, 379-394.

Guba, E. G., & Lincoln, Y. S. Fourth Generation Evaluation. Newbury Park, CA: Sage.

Gundry, K. & T. A. Heberlein (1984, Spring). Do public meetings represent the public? Journal of the American Planning Association, 50:175-182.

Hance, B. J., & Chess, C. (1994). Planning Successful Risk Communication: A training module for agency managers and staff. New Brunswick: Center for Environmental Communication

Heiman, M. (1990). From “not in my backyard!” to “not in anybody’s backyard!” Journal of the American Planning Association, 56: 359-362.

Kasperson, R.E. (1986). Six propositions for public participation and their relevance for risk communication. Risk Analysis, 6 (3):275-281.

Kihl, M. K. (1985). The viability of public meetings in transportation planning. Journal of Applied Behavioral Science, 21(2): 185-200.

232 Kraft, M. (1988). Evaluating technology through public participation: The nuclear waste disposal controversy. In M.E. Kraft and N.J. Vig, (Eds.), Technology and Politics (pp. 253-277). Durham: Duke University Press.

Laird, F. (1993, summer). Participatory analysis, democracy, and technological decision making. Science, Technology, and Human Values, pp. 341-36

Linney, J. A., & Wandersman, A. (1996). Empowering groups with evaluation skills: The Prevention Plus III Model. In D. M. Fetterman, S. J. Kafterian & A. Wandersman (Eds.), Empowerment evaluation: Knowledge and tools for self- assessment and responsibility (pp. 259-277). Thousand Oaks: Sage

Lynn, F., & Busenberg, G. J. (1995). Citizen advisory committees and environmental policy: What we know, what's left to discover. Risk Analysis, 15(2):147-16

Lynn, F. & Kartez, J. (1995) The redemption of citizen advisory committees: Perspectives from critical theory In O. Renn ,T. Webler & P. Wiedemann (Eds.), Fairness and competence in citizen participation: Evaluating models for environmental discourse (87-102) Dordrecht: Kluwer

Mazmanian, D. A., & Nienaber, J. (1979). Can organizations change? Washington: Brookings Institution.

National Research Council (1996). Understanding risk: Informing decision in a democratic society. (P. C. Stern & H. V. Fineberg, Eds.). Washington DC: National Academy Press.

Renn, O., Webler, T., & Wiedemann, P. (1995). A need for discourse on public participation: objectives and structure of the book. In O. Renn T. Webler & P. Wiedemann (Eds.), Fairness and competence in citizen participation: Evaluating models for environmental discourse (17-34) Dordrecht: Kluwer.

Renn, O.T., Webler, T. , Rakel, H., Dienel, P. & Johnson, B. (1993) Public participation in decisionmaking: A three-step procedure. Policy Sciences 26:189-21

Rosener, J. B. (1981). User-oriented evaluation: A new way to view citizen participation. Journal of Applied Behavioral Studies, 17(4), 583-596.

233 Rosener, J. B. (1982, July/August). Making bureaucrats responsive: A study of the impact of citizen participation and staff recommendations on regulatory decision making. Public Administration Review, pp. 339-345.

Scriven, M. (1997). Truth and Objectivity in Evaluation. In E. Chelimsky & W. Shadish (Eds.), Evaluation for the 21st Century (pp. 477-500). Thousand Oaks, California: Sage Publications, Inc.

Sinclair, M. (1977). The public hearing as a participatory device. In D. W. Sewell & J. Coppock (Eds.), Public participation in planning (pp. 105-123). London: John Wiley & Sons.

Slaton, C. D. (1992) Televote: expanding citizen participation in the quantum age. New York: PraegerYoung.

Tuler, S. and T. Webler. 1995. “A Process Evaluation Methodology.” Human Ecology Review 2 (Winter/Spring): 62-71.

U.S.Department of Defense, & Environmental Protection Agency. (1994). Restoration Advisory Board Workshop Guidebook. Washington.

U.S. Department of Energy. (1996). Site specific advisory board initiative: Evaluation survey results. USDOE.

Webler, T. (1995). Right discourse in public participation: An evaluative yardstick. In O. Renn T. Webler & P. Wiedemann (Eds.), Fairness and competence in citizen participation: Evaluating models for environmental discourse (17-34) Dordrecht: Kluwer

Webler, T. (1996). Organizing Public Participation: A Critical Review of Three Handbooks. Human Ecology Review. 3 (2):245.

Wengert, N. (1976). Citizen participation: Practice in search of a theory. In A. E. Utton, Sewell (Ed.), Natural resources for a democratic society (pp. 23-40). Boulder: Westview.

Wondelleck, J. M., & Yaffee, S. L. (1994). Building bridges across agency boundaries: In search of excellence in the United States Forest Service.

234 University of Michigan Ann Arbor: School of Natural Resources and Environment.

Young, C. W., G. Williams, and M. Goldberg (1993). Evaluating the effectiveness of public meetings and workshops: A new approach for improving DOE public involvement. Argonne, Illinois: Environmental Assessment and Information Division. Argonne National Laboratory. USDOE .

235

Elements of a Prudent Avoidance Policy

Tom McManus, Ph.D. Chief Technical Adviser, Department of Public Enterprise 44 Kildare Street, Dublin 2, Ireland Tel: +353 1 604 1023, Fax: +353 1 604 1189 E.mail: [email protected]

Introduction

I recall the first time I was asked about Ireland's EMF policy. It was April 1993. The location was Vancouver and the occasion was a workshop organised by Dr Kelly Gibney of BC Hydro. It was my first invitation to an EMF meeting as a participant. Until then I had been another anonymous body in the audience at Contractors Reviews and EPRI conferences. Now, at last, I was rubbing shoulders with the rich and famous, or at least the famous.

The workshop was opened by Dr. Gibney asking us to summarise our corporate or national EMF policies. Now policy statements by governments more often than not become the sticks their opponents use to beat them with. Ireland had no EMF policy then. Indeed, it has no formal EMF policy even now. So in response to Kelly's question I answered:

"Ireland's EMF policy is not to allow the EMF-health issue affect the growth and prosperity of the Irish economy".

Looking back I see that was more of a prayer than a policy. Yet prayer or policy, since 1993 the Irish economy has powered ahead with annual growth rates exceeding 8%. This year the forecast is for over 10% growth. The extra 2% is being attributed to our exports of sildenafil citrate of which Ireland is the world's sole producer. Those of you interested in finding out more about this remarkable chemical must ask their physician for a prescription for Viagra.

Economic growth in Ireland is closely related to energy consumption, and particularly to the consumption of electrical power. We have been successful, so far, in our handling of the EMF issue as it relates to power lines, to ensure that no new industry nor housing development is ever kept waiting for a power supply.

237 Were these new industries as dependent on mobile phone communications as electrical power the story would be different.

The subject of this paper is 'prudent avoidance' and how this policy is applied in Ireland to the EMF- health issue. In discussing this I shall concentrate on three main areas:

• ELF fields and powerlines - where we have done a lot and have kept problems largely under control.

• RF fields and mobile phone base station towers - where we are doing even more and the problems keep growing.

• Electrical hypersensitivity - where we are not sure what we should be doing, try many different approaches and find few of them work.

ELF Fields & Power Lines

Denver is where it all began. I searched for a suitable religious metaphor to describe the city. Not quite the Bethlehem of EMF, but undoubtedly a birthplace. Not quite the Mecca of EMF, but certainly a place to which we keep returning. The best I could come up with was EMF's Garden of Eden - where all the trouble started. Nineteen years later and one month ago I was in woodland near Cork tracking down a band of eco-warriors who are trying to stop the construction of a new 220 kV transmission line. Ten years ago my own involvement in EMF began when residents south of Dublin objected to a similar high voltage power line being built in their area.

So what has changed you might ask? One major significant change is the basis for the objections. Ten years ago it was EMF and health and childhood leukaemia. Today it is aesthetics. EMF appears to be taking a back seat. Cork has a magnificent deep - water harbour and is home to Ireland's greatest concentration of industry. Power is supplied to this industry by overhead power lines. Industrial expansion - including a $½ billion expansion to the Pfizer plant, to make more of you know what - means more power lines and more pylons. Some among the local population are saying enough is enough.

Essentially the public are displaying an attitude with which we are all familiar. People are happy to enjoy the benefits of science - in this case, electricity;

238 they just don't want the technology that produces these benefits. A particular difficulty faced by those in Ireland who object to a development on aesthetic grounds is that under Irish law no one has a right to a view. Although the Cork transmission line has passed all legal obstacles, the electrical utility may have to serve injunctions to remove trespassers from the right of way before construction can proceed.

The evolution of our approach to handling power line problems was discussed at the WHO meeting in Vienna, (McManus, 1997). This paper is confined to discussing what we do today.

New power lines

As far as is practical the centreline of a new high voltage (220 kV+) power line will be no closer than 50 metres to a farm, factory or residential building. In any event the line will not come closer than 23 metres (25 yards in Imperial measure). Double circuit lines are now always asymmetrically phased to minimise magnetic field levels.

The electricity company submits its application for permission to build the new line to each of the counties (the planning authorities) through which the line passes. The county councils will seek a report from the Dept. of Public Enterprise which will review all recent developments relevant to the power lines EMF - health issue. That review will encompass the results of the latest epidemiological and animal exposure studies, the current views and opinions of overseas national and international health advisory groups, the prevailing exposure standards and guidelines issued by ICNIRP and various national agencies, and the actions and decisions being taken by governments in other countries relating to similar developments.

The Department of Public Enterprise endeavours to hold the high middle ground on EMF by providing to all parties who seek it, including objectors, information that is balanced, fair, accurate and up-to-date.

New housing

In recent years Ireland has been building nearly 40,000 new houses a year. This is a substantial number for a country with a population of only 3½ million. Many of these new homes are encroaching on existing overhead transmission lines

239 and a good proportion of the prospective owners in such cases will contact the Dept. of Public Enterprise. Their concern is invariably a worry over exposure to electromagnetic fields from the adjacent power line. We will receive an average of twenty such calls a month.

Our response to these enquiries has three main elements:

• First, we will spend up to 30 minutes on the phone discussing the caller's concerns and answering their questions.

• Secondly, we will send them that day a standard letter which is updated regularly and which summarises the Department's position concerning powerline fields. With this letter we will also enclose a wide selection of recent papers, statements and reports from around the world.

• Thirdly, we will organise a survey for the caller. This survey will involve measuring electromagnetic fields at the new home, between the new home and the power line, and - this is very important - in the caller's present home. The surveys are undertaken usually within a few days of the call and even by the following day, if a decision on purchasing a new property cannot be delayed. The surveys are carried out by consultants to the electrical utility and at no cost to the caller or to the Department. The caller or a representative is encouraged to accompany the surveyors on their rounds.

In over 95% of cases we find the callers well satisfied with our response. Indeed the openness with which we provide information and the speed with which we can carry out surveys goes a long way to reassuring the caller even before the survey has taken place.

We always make the point in our standard letter that domestic exposures to ELF are invariably well within the ICNIRP guidelines. However the main reassurance, I believe, comes from the individuals being able to compare the fields in their old home, which is associated with comfort, safety and normality, with fields in the new home. These fields are usually quite similar. Often the field strength between the house and the power line will pass through a minimum value. This provides further reassurance. There have been only a handful of cases where a caller decides not to progress the purchase of a new home on the basis of its proximity to overhead powerlines.

240 Phone Masts

Mobile phone ownership in Ireland is growing at a rate of 40% a year. Likewise the number of cellsites is growing at a comparable rate. The popularity of the phones is equalled only by the dislike of the base station transmission towers which are springing up everywhere. Opposition to these towers is widespread. The main fear is that the towers will cause cancer among those who live nearby. The risk to children's health is often advanced by objectors to these developments. The Dept. of Public Enterprise can receive as many as a dozen calls in one day from concerned members of the public over the risks posed by the fields from the phone towers. It is however rare to have anyone express concern about the phones themselves which expose their users to fields 1000 times stronger.

Because of the inadequacy of the public telephone system outside the major cities there are powerful social, economic, and security reasons for having a first class mobile phone service. The government has licensed two carriers to provide a service on the 900 MHz band and a third carrier is in the process of being awarded a licence for an 1800 MHz service. However, the development of these services is being frustrated by public opposition, orchestrated by environmental activists. These groups apply pressure on elected representatives and senior officials at county level to have the planning applications for towers refused. Needless to say it is in those parts of Ireland most in need of a mobile telephone service that the problems are greatest and the development of the service most delayed. We have launched a number of initiatives with the objective of assisting the planning authorities and informing the public.

In July 1996 our Department of the Environment with the co-operation of the Departments of Health and Public Enterprise produced guidelines concerning the siting of phone towers (Environment, 1996). Among the recommendations was one that new towers should be located in residential areas only as a last resort or where there were strong technical reasons which ruled out an alternative site. Today, however, it is common for a county council to refuse planning permission on the grounds that the proposed development is in a residential area and so contravenes the official guidelines. The county councils who must give a reason for rejecting a planning application have chosen to elevate a guideline to a mandatory requirement and ignore the qualifications associated with the original guideline.

In March 1998 the same three government departments organised an international conference specifically to address the phone mast-health issue (Public

241 Enterprise, 1998). Our list of speakers read like a Who's Who of EMF. Our capacity audience of around 350 comprised all shades of opinion from phone mast objectors groups to the phone companies themselves. It included over 40 elected representatives (politicians) and officials from nearly every planning authority in the country. By the end of the conference it was felt the health issue had been put on the back burner and into a realistic perspective. However it was very clear there was major public antipathy towards the phone companies themselves, who were seen as arrogant, uncommunicative, and eager to exploit every loophole in the planning laws to erect towers with a minimum of public consultation.

In the months that followed the March 1998 conference the Parliamentary Joint Committee on Transport and Public Enterprise - this is analogous to a Congressional Sub-Committee in the US - held two meetings on the mobile phone tower issue. At these meetings the Secretarys General of Public Enterprise, Health, and Environment; the chief executives of the phone companies; the Director of Telecommunications Regulation, and many other senior officials were all closely questioned. Representatives of a dozen anti-phone tower groups were also given an opportunity to put their concerns to the Joint Committee. The Committee's report is in preparation (Joint Committee, 1998), but I believe it will make recommendations concerning the strengthening of our Siting Guidelines document and encourage greater sharing of towers by the phone companies for their antennae. It is also likely they will call for improvements in the monitoring of field strength levels around base station towers.

In June 1998 the Telecommunications Regulator published the results of the first nation-wide survey of radiofrequency emissions (ODTR, 1998). The survey involved visiting a sample of 30 broadcasting antennae. The sample included mobile phone transmitters, radio transmitters, television transmitters and microwave communication transmitters. At each site a search was made for the location at ground level where the public had access and where signals were strongest. Signal strengths were then measured and the measurements compared to the ICNIRP guidelines. Apart from two medium wave radio transmitters which were inadequately fenced all measurements fell within the ICNIRP guidelines by orders of magnitude.

However despite all of these efforts many Irish planning authorities, bowing to public protest, continue to refuse planning consent to new phone towers. It could be asked where are we going wrong?

242 In our handling of phone mast enquiries to the Department of Public Enterprise our approach is similar in some respects to that described for new houses near pylons. We discuss the problem, answer the questions, send out a standard letter that reflects the Department's current view, and include a selection of recent literature and publications on the subject. However there is one thing we are unable to do. We cannot offer a monitoring service as a matter of routine as we are able to do with powerlines. If we wish to carry out a survey we must engage consultants and pay for them ourselves at a cost of approx. $1500/survey. Our limited budget restricts us to no more than three or four such surveys per year. The kind of RF survey we commission is different from that undertaken by the Telecommunications Regulator. We will go the caller's home and measure all signals in the frequency range 30 MHz to 1000 MHz. We will extend the frequency range to 3000 MHz in future surveys when new equipment becomes available.

Typically a survey will identify some 50 or so strong signals. These are ones that can be readily picked up on a good radio or TV set. The signals of course include those broadcast from the analogue and GSM phone towers. We then identify the sources of the main signals and express them in order of strength in units of milliwatts per square metre. We don't use decibels. The survey compares the signal strengths from phone towers to those from VHF radio and UHF television transmitters. From this data it becomes quite clear to the householder that the signals from the phone tower are similar in strength to those coming from their favourite radio and TV stations and often much weaker. We do, of course, relate the measured signals to the ICNIRP guidelines, but it is the comparison among the actual signals themselves that provides the greater reassurance to the householder.

The Department has held discussions with the phone companies with a view to setting up this kind of monitoring as a routine service but have so far failed to convince them of its merit.

Electro-sensitive Individuals

There are not many electro-sensitive people in Ireland, but they are quite vocal and have their own pressure group. Generally the symptoms they experience include unexplained rashes and skin troubles. Many also suffer chronic fatigue syndrome. They are certainly not inventing their problems. They are genuinely suffering. They have usually exhausted the resources and patience of conventional medicine and have then sought help from alternative medicine. When the solutions proferred by alternative medicine have also failed, their problems must therefore be

243 due to the only thing left to blame - electromagnetic fields - which is where the Dept. of Public Enterprise comes in.

We listen sympathetically. We remind callers that despite the huge advances in medicine the medical profession can still only diagnose about half the illnesses presented to them. Fortunately most of us get better on our own. A few unfortunate people never get better. This nostrum does not always help. People demand good health as a right. One lady I spoke to stated, and I quote:

"It is the governments' responsibility to find out what's making me ill and to stop it."

The remit of the Department of Public Enterprise covers gas, electricity, oil, solid fuel, renewable energy, public transport, telecommunications, and ionising radiation. We will send teams round to check for gas leaks, carbon monoxide, electricity faults, radon concentrations, and EMF levels at both ELF and radio frequencies. Usually everything is perfectly normal. We will check the water for lead if we are in an older part of town. We will direct the electro-sensitive to a self- help group started in Sweden. Should we be told there is a cluster of similar illness in the neighbourhood and this can be substantiated we will contact the local Health Authority who will send public health specialists to talk to the people concerned.

Why do we get involved? Why do we bother? It's not entirely altruism. The electro-sensitive are well organised and have good political contacts. In many respects Ireland is still like a large village. Parliamentary time can be taken up by an elected member enquiring of a Minister why Mrs. Murphy didn't get her pension last week. Parliamentary questions generate action. Likewise do letters to Ministers, especially letters from opposition members writing on behalf of individuals who claim electromagnetic fields are damaging their health and accusing the government of doing nothing about it. We at least enable our Minister to respond that she has or is taking action.

So far we have failed to identify any abnormal electromagnetic field strengths in the environment of the electro-sensitive. Sometimes we might even be thanked for at least listening and doing something. But most remain totally convinced of the cause of their problem and it is not uncommon for them to be invited to promote their views on radio talk shows. We have had a few successes. Once we 'cured' someone who was electro-sensitve. This was when we had a gas

244 leak repaired, had proper ventilation installed - at the state's expense, and persuaded a volunteer group to give the house its first real clean in ten years.

Prudent Avoidance

Now it is time to say something about prudent avoidance.

The merits and demerits of prudent avoidance have raged since Granger Morgan (1988) first proposed the policy ten years ago. Prudent avoidance essentially involves accepting the principle that while there is no firm evidence to conclude electromagnetic fields present a health risk under normal circumstances, one should reduce one's exposure to these fields if this can be done with a minimum of effort and expense. Prudent avoidance in the home, for example, can involve moving an electric clock to the far side of a bedroom. To some, prudent avoidance is akin to walking around a ladder rather than under it. It is superstition not science.

There is also an important ethical question associated with prudent avoidance. Resources are finite and in allocating limited resources to public health, governments must always consider the greater good for society. It could be argued that it is hardly ethical to commit resources to avoid a theoretical health effect when governments lack the funds to tackle real actual health problems.

On the other hand governments are continually faced with decisions they have to make and actions they have to take in the absence of all the facts and before science has provided a definitive answer. In such cases the policy adopted will often be "The policy of least regret". The policy of least regret can be expressed as the choice of two alternatives:

1. What is the possible penalty if we do nothing and the worst happens?

2. How much money will be wasted if we take some action and the problem never arises?

It's the same question most of us face each year when our home insurance policy comes up for renewal. Prudent avoidance sits well with the policy of least regret. One country that has adopted the policy of prudent avoidance and supports it with an excellent rationale is Australia.

The Gibbs (1991) report from New South Wales in 1991 stated:

245 "It has not been established that electric fields or magnetic fields of power frequency are harmful to human health, but since there is some evidence they may do harm, a policy of prudent avoidance is recommended."

Two years later the Peach Panel report to the Victorian government (Peach, 1993) added its endorsement of prudent avoidance which it saw as a means of:

"... looking systematically for strategies which can restrict field exposure and adopting these strategies which seem to be prudent investments, given their costs and the level of scientific understanding about possible risks."

In reaching this conclusion the Panel recognised:

"... that a policy of prudent avoidance is not a health-based policy, and that the implementation of the policy cannot be seen as necessarily being of benefit to public health, and to any individual."

.....which is a polite way of saying it might all be a waste of time and money.

In December 1995 an Australian Senate Committee (Australia, 1995) again endorsed current Australian policy concerning prudent avoidance:

"The Committee agrees that, as a minimum policy, or until evidence suggests otherwise, the concept of 'prudent avoidance' should continue to be practised by government and power authorities [and] in the case of [the powerline project] prudent avoidance should mean siting the line as far as possible from houses, outbuildings and other farm facilities."

In Ireland new electric powerlines are routed in accordance with similar principles. Double-circuit lines are asymmetrically phased to minimise magnetic field strengths. The electrical utility and the Dept. of Public Enterprise also provide public information on minimising personal exposure to electromagnetic fields.

In the case of mobile telephone communications, public exposure to the radio frequency field is an essential element of the technology. There are no

246 appropriate or relevant prudent avoidance measures available to the public even if they were considered necessary or prudent. The popularity of the mobile phone is increasing the number of new cells required. More cells in operation leads directly to a reduction in signal strengths from both the base station transmitter and from the mobile phone. When eventually everyone has a mobile phone, public exposure to the associated radiofrequency fields will be at a minimum.

The Future

Handling the EMF-health issue in Ireland is likely to become more difficult in the future. Three factors influencing this are:

• Affluence

• Litigation

• The forthcoming NIEHS report.

Affluence

Ireland is now an affluent country with a standard of living now above the European average. And affluence creates hypochondria. Affluent people feel entitled to good health as a right. Any threat to good health, however obscure, remote or unlikely, will generate loud demands that the authorities do something about it. "Convince us that this technology will do no harm!" say the activists. "You can't prove a negative." answer the health professionals. I now see this familiar exchange being employed in reverse. If one accepts that you can't prove a negative then logically anything might be true. There is therefore scope for quite outrageous and extreme claims to further strike fear among the public.

So why should people in Ireland and in the west generally who are among the healthiest and longest-lived in history be so concerned about their health? At a Harvard Club Conference (Harvard, 1991) on this topic the panic was attributed to:

• those promoting the concerns who are moved by self-interest rather than facts,

• the presentation of problems by a media that is less interested in science than in sensationalist sound bites for the public,

247 • the real difficulties of dealing with the concept of risk.

Another factor is that we now live in an age where many feel that the secret of everlasting life on earth is just around the corner or at least the prospects of a greatly improved and vigorous life-expectancy. In these circumstances one shouldn't be surprised to find the public anxious to avoid any exposure that might prevent them being around when the anti-ageing pill comes on the market.

For those in the front line, dealing with the public on EMF issues will not become easier. I recently came across a reference to the opening of the first international air service from London to Paris in 1920. One of the main fears of the public at the time was the risk to life of standing up on open-topped buses. One can only wonder if, in 70 years time, some present concerns over electromagnetic fields will seem just as bizarre.

Litigation

A feature of civil actions under US Law is that the loser does not have to pay the winner's costs. This, and the fact that the US system allows the plaintiff's lawyers to work on a contingency fee basis, provides lawyers with strong incentives to engage in speculative litigation.

In the United States it proves cheaper for a company like Dow-Corning to settle claims than defend itself, even though in the case in question - breast implants - there is no evidence that the incidence of ill-health among women with breast implants is greater than in the wider population. There is a danger that something similar could happen with mobile phone users or with people living near power lines.

A single lost or undefended EMF case in the United States would be exploited by anti-EMF groups in Ireland to seek enhanced exposure restrictions and to prevent the growth of the electricity and telecommunications industries.

Ireland's and possibly Europe's first case where an injunction was sought against the construction of a base station transmitter on the grounds of a serious and immediate hazard to health was heard in Dublin's High Court earlier this year. The request for the injunction was brought by under-age children, represented by their parents. Such an approach is aimed at generating public sympathy and discouraging

248 the defendants from seeking costs should the action fail. Although the plaintiffs were supported by a huge weight of evidence (I speak in gravimetric terms) submitted by their technical experts, the judge (High Court, 1988) refused the request, saying:

"On the evidence before me, I have come to the conclusion that it is highly improbable at the very least that any injury will ensue to the children between now and the hearing of the action and I think that that is as far as I should go in terms of comment on the facts. In those circumstances, I do think that it would be just as far as the Defendant is concerned, as not only would it be prevented from going ahead with the Easkey project but in reality the injunction would have very considerable repercussive affects likely to damage the Defendant's highly competitive business throughout the country.

For the reasons which I have indicated early on in this judgement, I am very reluctant to adopt the traditional language of "serious issue to be tried, balance of convenience and damages as an adequate remedy" and I believe that this is a case in which those categories are neither appropriate nor helpful. But in case I am wrong about that, I would still have to come down in favour of refusing an injunction. Having regard to the stringent requirement of probability before a quia timet injunction would be granted on a permanent basis, I am very doubtful that there is a serious issue to be tried here, but even if there is, I have no doubt that for the reasons I have indicated, the balance of convenience requires that I should refuse any temporary injunction pending the hearing of the case.

I therefore refuse the interlocutory relief sought."

A full hearing of this case is expected to take place next year. It is my belief that this will be the first of many such cases, especially if anti-EMF groups can win an action in the United States.

NIEHS report

Should the final NEIHS report to Congress declare powerline frequency fields to be a possible human carcinogen this conclusion can be expected to receive

249 major media coverage in Ireland. In one respect NIEHS would simply be restating the position as it was prior to the RAPID Program, for had there been no possibility of EMF being a human carcinogen there would have been no need for RAPID or any other of the major research projects underway around the world. So what has changed?

My personal view is, that should NIEHS endorse the statement in its draft report (NIEHS, 1998), authorities may well conclude: "If we can't get a definitive answer to this question after spending $46 million on RAPID then we're never going to get a definitive answer. And we're certainly not going to put another $46 million into finding out!" Authorities will then be in their familiar position of making a decision on the basis of incomplete and uncertain information. My forecast is that they will lean towards recommending a policy of "prudent avoidance" and may set up committees to identify specific actions and requirements. Other countries would then consider whether to follow the US lead. I would also predict that EMF lawyers would endeavour to show that their clients' illnesses were due to being exposed to fields from installations that did not meet these specific requirements.

References

Australia, 1995: Eastlink : The Interconnection of NSW and Queensland Electricity Grids with a High Voltage Power Line. Senate Committee Report, Parliament of the Commonwealth of Australia, December 1995.

Environment, 1996: Telecommunications Antennae and Support Structures - Guidelines for Planning Authorities. Dept of Environment, July 1996.

Gibbs, 1991: Inquiry into Community Needs and High Voltage Transmission Line Development, The Rt. Hon. Sir Harry Gibbs, Report to New South Wales Government, February 1991

Harvard, 1991: Health, Lifestyle & Environment, P. Berger et al, The Social Affairs Institute and The Manhattan Institute, 1991.

High Court, 1998: Szabo et al v. Esat Digifone, Ireland et al., Judgement of Mr. Justice Geoghegan, delivered 6 February 1998.

250 Joint Committee, 1998: Report on Non-Ionising Radiation Emissions from Communications Masts. Joint Committee on Public Enterprise and Transport, (in publication, 1998).

McManus, 1997: Practical Implementation of EMF Communication, T. McManus, in "Risk Perception, Risk Communication and its Application to EMF Exposure". ed. R. Matthes et al, ICNIRP 5/98 International Commission on Non-Ionising Radiation Protection, 1998.

Morgan, 1988: Controlling Exposure to Transmission Line Electromagnetic Fields : A Regulatory Approach that is Compatible with the Available Science, M.G. Morgan et al. Public Utilities Fortnightly, 17 March 1988.

NIEHS, 1998: Assessment of Health Effects From Exposure to Power - Line Frequency Electric and Magnetic Fields. Working Group Report. NIEHS, July 1998.

ODTR,1998: Compliance with Emission Limits for Non-Ionising Radiation. Office of the Director of Telecommunications Regulation, July 1998.

Peach, 1993: Report of the Panel on Electromagnetic Fields and Health, Prof. Hedley G. Peach, Report to Government of Victoria, September 1992 (published 1993).

Public Enterprise, 1998: Communications Technology and the Community. Proceedings of March 1998 Conference, Dept. of Public Enterprise, July 1998.

251

The application of prudent avoidance in EMF risk management

Michael Dolan Deputy Director & General Counsel, Manager, EMF Advisory Group Electricity Supply Association of Australia Limited GPO Box 1823Q, Melbourne, VIC 3001, Australia 3000 Tel: +61 3 9670 1017, Fax: +61 3 9670 1105 E-mail: [email protected]

and

Kev Nuttall, Principal Network Investigations Engineer, ENERGEX Paul Flanagan, Manager Environmental Affairs, Pacific Power Garry Melik, EMF Consultant, Magshield Products International Pty Ltd.

Abstract

More than twenty years of international scientific research has not yet resolved the issue of whether or not exposure to power frequency (50/60 Hz) electric and magnetic fields causes adverse health effects such as cancer.

Scientific research is continuing, but it may be some time before the issue is resolved. In the meantime, governments, regulatory bodies and the electricity supply business must manage both the issue itself as well as the public perception of it.

In Australia, the electricity supply business has a long-standing policy of acting prudently in relation to the issue. Acting prudently includes practising prudent avoidance when building new transmission and distribution facilities. This policy position has been endorsed by Federal and State Governments and is an integral part of managing the issue.

This paper outlines the background to the adoption of the policy of prudence in Australia and gives practical guidance as to the way in which prudent avoidance may be applied by electricity supply organisations when designing and building new transmission and distribution systems. The paper argues that prudent avoidance should be an essential part of EMF risk management.

1. Introduction To EMF Health Issue

We are all aware of concerns amongst members of the general public regarding the possibility of adverse health effects such as cancer from exposure to power frequency electric and magnetic fields (EMFs). In recent years, scientific research has focussed on magnetic fields rather than electric fields, and the remainder of this paper reflects that focus.

253 Scientific discussion regarding the issue dates back to the late 1960’s with the introduction of ultra high voltage transmission systems. Concerns were raised that the high electric fields associated with the operations of such systems might have health implications for electricity utility workers. In 1972, general health effects such as headache and fatigue were reported in switchyard workers exposed to high electric fields in the former Soviet Union. Subsequent studies attributed these symptoms to aspects of the work environment other than EMFs.

Public concern was heightened by an American study in 1979 which suggested a link between exposure to estimated magnetic fields from street distribution lines and transformers and childhood cancer. This was a turning point in the issue as it was the beginning of the transition from electric fields to magnetic fields as the source of concern. There has been much research since that time. The scientific literature is both complex and voluminous and covers a wide spectrum of science including epidemiology, in vitro and in vivo laboratory research, and engineering.

All of the research has been extensively reviewed by Australian and international inquiries and expert panels established for the purpose of trying to determine whether or not human exposure to EMFs is related to adverse health effects. There is scientific consensus that health effects have not been established, but that the possibility cannot be ruled out. Many scientists agree that there is a need for ongoing high quality scientific research in order to try and give better answers to the questions which have been raised.

2. The Gibbs Report

On 28 May 1990 the Minister for Minerals and Energy for the State of New South Wales (“the Minister”) authorised former Chief Justice of Australia The Right Honourable Sir Harry Gibbs G.C.M.G., A.C., K.B.E. to conduct an inquiry into community needs and high voltage transmission line development in the State of New South Wales. On 3 July 1990 the Minister wrote to Sir Harry Gibbs a letter which included the following paragraph:

“Without in any way limiting or restricting the nature of the terms of reference of your Inquiry, I would like to request that you specifically include in your investigations the question of electromagnetic fields and their relationship to health.” [1.1.4]

254 The Gibbs Report was delivered to the New South Wales Minister for Minerals and Energy in February 1991. In making findings relating to the state of science regarding the health issue, former Australian Chief Justice Sir Harry Gibbs was assisted by four Australian scientists. The Report concluded:

• “It has not been scientifically established that electric fields or magnetic fields created by the electric power system in New South Wales (or by any electric fields or magnetic fields of extremely low frequency) initiate or promote cancer or have any other harmful effect on humans. However, it has not been scientifically established that such fields are not harmful”.[5.11.1]

• “It would serve no useful purpose to review all of the numerous studies done “in vitro” or “in vivo” with a view to attempting to discover the effect of (the fields) on biology or health. It is quite apparent that the studies do no more than show that the fields can cause biological changes; they certainly do not enable it to be concluded that the fields are harmful to health.” [5.4.14]

• “The epidemiological studies into the effects, if any, of electromagnetic fields at extremely low frequencies support the view that it is possible that exposure to those electromagnetic fields causes an increased risk of developing cancer in childhood and an increased risk in adults of developing leukaemia, lymphoma and brain tumours. The view that this is possible is particularly supported in the case of children by the residential studies and in the case of adults by both the residential and occupational studies. However, the findings are by no means conclusive and it cannot be said categorically either that it is probable that there is a risk, or that there is no risk.” [5.5.6]

• “If any risk is created by exposure to extremely low frequency electric fields or magnetic fields, there is no reason to suppose that it is created entirely, or even mainly, by exposure to the fields created by high voltage transmission lines. Electric fields and magnetic fields are produced also by distribution lines, by wiring in homes, shops, offices and schools, by return currents in underground gas pipes and water pipes and in the ground itself, and by the many appliances used in the home, at work, and in hospitals.” [5.11.3]

255 • “The fact that evidence on the question whether exposure creates a risk to health is so inconclusive suggests that if a risk exists it is a comparatively small one...... ” [5.11.5]

More research has been published since the Gibbs Report, and many reviews by other independent and authoritative panels have been undertaken. It is suggested that no more definite conclusions have been able to be drawn from this expanded body of literature, and that the various findings and conclusions reached in the Gibbs Report in 1991 are as relevant today as they were at the time they were published, and provide a sound basis for public policy decision making.

3. Emf Policy Issues

3.1 Background

Any policy or regulatory actions proposed regarding EMFs should be based on the totality of the science and not on the results of a single research project or the opinion of any single scientist irrespective of his or her standing. The considered conclusions of independent authoritative reviews of the body of science by groups of expert scientists or official inquiries or panels established for the purpose are an essential guide for health authorities and other decision makers when dealing with the issue.

The uncertain state of the science regarding EMFs has presented significant challenges for decisionmakers. As a Texas PUC Review put it:

“Policy Issues and Options

The current status of scientific evidence regarding EMF health effects is unclear. There is no definitive indication that EMF exposure can affect health, and there are no data that establish convincingly that it does not. In fact, as is often the case in situations involving very low probability cause/effect relationships, it may not ever be possible to prove an effect or the lack of an effect.

With respect to the EMF health effects issue, state legislatures find themselves in a quandary. Acceptance of false positive conclusions may result in a significant expenditure of taxpayers’ money and divert attention from efforts to seek the true source of any increased risk. By contrast, not

256 acting on false negative conclusions is likely to be interpreted by the public as irresponsible disregard for citizens’ safety. Therefore, it seems reasonable to expect legislatures to actively support efforts to resolve the conflict.

Regulatory agencies normally address scientific uncertainty, such as the EMF health effects question, through procedural mechanisms similar to those used in the courts and legislatures. The details of the mechanisms vary considerably depending on the nature of the regulatory agency and its legislative charter. Political pressures to “do something” about the EMF issue may result directly or indirectly in the search for regulatory relief, especially if no action is achieved at the judicial or legislative levels.

In at least 17 states, legislative or administrative agencies have formally considered the possibility of health effects as a result of EMF exposure. Responses range from dismissal of the question due to lack of evidence (Wyoming) to codification of formal EMF limits in transmission lines (Florida). Courts and legislatures are actively considering actions in several states.

Different responses and their rationales are tied to different views of what constitutes the key problem in the EMF debate. There have been at least four different ways to define the EMF “problem”, each with distinctive views of the scientific evidence, of the proper role for science to play, and of the proper perception of risk. More importantly, each definition carries a policy prescription along with it. In the absence of a conclusive body of scientific findings that would provide a firm grounding for deciding which of the four ways of constructing the problem is the most appropriate, one is left to decide largely on the basis of pre-existing beliefs and values that each of us brings to the EMF issue.

In this instance, the values of experts alone may provide too narrow a basis for legitimating one definition of the problem over others. Recognising this limitation, the Committee recommends that, until science can provide a clearer path, state officials should engage the public in open discussions of both the evidence to date and the public values that influence its interpretation.”

257 Professor Granger Morgan of Carnegie Mellon University in the United States has written widely on the issue of public policy in the context of EMFs. As he wrote in an essay for a book (Carpenter & Ayrapetyan, 1994) on the issue:

“In the face of possible risk and incomplete evidence many (including myself) believe that the wise strategy is to exercise prudence. Search for low-cost steps that could get people out of fields, especially new fields, and then adopt them. At the same time, do not go off the deep end. On a per capita basis, do not spend large amounts of money or incur great inconvenience until the evidence is clearer than it now is.

In our private lives when we face a possible but uncertain risk, we exercise such prudence all the time. For example, many of us eat a bit more fibre, broccoli, and cauliflower and a bit less fat and charbroiled meat than we once did. We know that these dietary changes offer no assurance of protection from cancer, but there is some evidence that they might help. The changes do not cost much. You can get used to them. So, it seems prudent to make some adjustments. In public life we have rather greater difficulty exercising prudence. Our regulatory and legal systems want to classify everything unambiguously as either safe or hazardous. Indeed, a U.S. Federal Appeals Court judge once told me, “no matter how incomplete the scientific evidence, present me with the case and I will decide”.

Creating certainty by fiat in the face of serious scientific uncertainty is not a wise way to organize a society. Understanding this, a variety of groups and individuals have begun to develop strategies for exercising prudence on the subject of power-frequency fields.”

The Gibbs Report dealt with these policy aspects of the EMF health issue as follows:

• “It then becomes a question of policy what action should be taken to avert a possible risk to public health when it cannot be said either that it is probable that the risk exists or in what circumstances a risk, if one exists, arises. A suggestion has been made in the United States that a policy of prudent avoidance should be adopted.” [5.11.6]

• “It would not be prudent, but foolish, to make radical or expensive changes to existing lines until further scientific studies have resolved the

258 doubts. On the other hand, when new lines are being constructed, it may be prudent to do whatever can be done without undue inconvenience and at modest expense to avert the possible risk, remembering that if that is not done and future research establishes the existence of a real risk to health, serious problems may arise which can be remedied only at great cost.” [5.11.9]

The policy recommendation contained in the Gibbs Report formed the basis of the policy subsequently adopted by the Australian electricity supply industry.

3.2 Adoption of policy prudence by ESAA

At a meeting in Sydney on 19 July 1991, the Board of Directors of ESAA adopted a formal policy in relation to electric and magnetic fields. The ESAA policy on EMF was adopted in the light of authoritative reviews having concluded that no adverse health effects from exposure to EMF had been established, but recognising that there was, within the Australian community, some genuine public concern about the issue which had to be addressed by the industry. This policy is reviewed at each meeting of an EMF Policy Committee which normally meets at six monthly intervals.

The policy recommends to ESAA members, including distribution members, that they operate their electrical power systems prudently within Australian health guidelines, and closely monitor, and, where appropriate, sponsor high quality scientific research. The policy is attached to this paper as Appendix 1.

ESAA committed itself to ongoing public and workforce education on the issue. This has included the sharing of information and the undertaking of measurements by ESAA members where requested. The industry also became involved in supporting Australian EMF research. In November 1993, ESAA commissioned a $1 million laboratory study examining the role, if any, played by exposure to magnetic fields in the cancer process in mice. The results of this study were published in March 1998. The study results were negative. ESAA has also supported other EMF research.

In an accompanying Advice to its members, the ESAA Board clarified what it meant by acting prudently in this context. The document was amended by the Board in June 1992, November 1994 and again in March 1997. In ESAA’s view, acting prudently means embracing a range of sensible actions having regard to the

259 uncertain state of the science, and which take into account scientific research and community concerns. These actions are set out in the document, and include informing employees and the public about the issue, and practising prudent avoidance (as defined in the Gibbs Report) when designing and building new transmission and distribution facilities. Such actions can include considering the design of the new facilities with respect to the EMFs which may be produced, sharing information on EMFs with the community, and taking community views into account when siting new facilities.

In its formal parliamentary response to the December 1995 Senate Eastlink Report the Australian Federal Government endorsed a policy of prudent avoidance for the building of new transmission lines in Australia. In a document tabled in the Parliament in October 1996 the Federal Government expressed its support for ESAA’s policy of prudence including the practice of prudent avoidance.

The Federal Government noted that, had the Eastlink project proceeded - it was cancelled in 1996 following a change of government - prudent avoidance would have been included as part of the normal operational procedures of the two transmission authorities Powerlink and TransGrid. This would have involved the transmission line being constructed well away from houses, schools and similar structures.

In its response the Federal Government noted that “the potential health impacts of electric and magnetic fields (EMFs) have caused considerable concern within the communities potentially affected by the Eastlink proposal. This could reflect the situation that even with major continuing scientific research programs at the global level, a clear link between EMFs and health problems has been neither established nor disproved.”

The Federal Government also acknowledged “that the appropriateness of prudent avoidance as a national public health policy needs to be assessed on the basis of the understanding of potential health effects at any given time.”

4. Prudent Avoidance Principles

In Australia there is broad consensus that a prudent approach should be taken in the design and siting of new transmission facilities. Although there is no precise definition of prudent avoidance, there is considerable discussion in the

260 literature which provides guidance as to how it might be applied in practice. In particular, Sir Harry Gibbs described prudent avoidance as:

“....doing whatever can be done at modest cost and without undue inconvenience to avoid the possible risk (to health)”.....”

This raises the question as to what might constitute “modest cost”. In 1993, the California Public Utilities Commission in the United State of America published an order defining prudent avoidance as undertaking suitable activities up to 4% of the cost of a new electricity company installation project. In a guidance document produced subsequently by Southern California Edison Company, the recommended approach was to:

“Implement reasonable no cost and low cost steps to build new electric utility lines and substations in ways that reduce magnetic fields.”

The application of prudent avoidance in the design and construction of new electrical facilities is a process of assessing the extent to which people may be exposed to fields produced by them and considering what “low cost” and “no cost” measures might be taken to reduce such exposure within acceptable constraints. Other than in California we are not aware of any regulatory order or direction dealing with the quantum of cost.

In broad terms, the range of measures which may be available to reduce exposure to the magnetic fields generated by electricity supply business facilities come under two broad generic headings:

1. siting measures

2. design measures.

4.1 Siting Measures

The process of site selection of electricity supply business facilities is a complex one involving a multitude of considerations of which the possible adverse health effects of exposure to EMFs is but one. Other considerations include:

• the location of the power source and the load

261 • provision for future development

• the location of existing rights of way

• cost considerations

• the nature of terrain and other siting constraints such as dams

• residential areas

• airfields

• national parks

• sites of particular cultural or heritage value

• transport corridors and the like.

In the context of the siting of electricity supply business facilities, the issue of EMFs is rarely an overriding consideration but rather should be considered as one of several important factors.

Furthermore, because many of the factors which influence the siting of electricity supply business infrastructure are, at least to some extent, of a sociological nature, an essential part of the siting process should be the engagement of the affected community in the process of siting. This requires the community to be informed of the proposed project at an early stage, acquainted with the range of factors which may be relevant to the siting decisions and, genuine input sought. In respect of EMFs, the community involvement process could include measures such as informing the community about the need for the facility and the various site selection constraints, providing educational material (preferably including material from independent sources on the issue of EMFs), providing factual information on the magnitude and extent of fields likely to be associated with the proposed facility, providing information regarding the magnitude and extent of EMFs in the general area and in typical everyday situations, eg. in the home, the street, etc.

The community should be informed at an early stage of the consultation process to be followed and the way in which their views can be fed into it.

262 Community feedback should be factored into project planning, along with the various other factors, and a decision made. Experience in Australia with community consultation programs suggests that dwellings, schools, playgrounds and other locations, especially those frequented by children, are likely to be of most interest from the perspective of EMFs. Where these can be avoided at modest cost and minimum incovenience it is prudent to do so.

4.2 Design Measures

Most of the design measures that will be discussed in the remaining sections of this paper are presented in the EMF mitigation handbook written by G Melik in 1996 under contract to the Electricity Supply Association of Australia Limited.

The following design measures for reducing magnetic fields may be applied to power lines of all voltages:

• Increasing the (vertical) distance of the line from sensitive receptors

• Configuring the conductors to minimise the magnetic field

• Arranging the phases to minimise the magnetic field

• Using more than one conductor per phase (split phase) and arranging them to minimise the magnetic field

• Reducing the current

• Shielding or active cancellation

• Locating the lines underground (in some cases this can increase the ground level magnetic field but the field strength will diminish more rapidly with distance).

The option selected should neither jeopardise the reliability nor downgrade the operating characteristics of the electricity system. Nor should it create a hazard to maintenance personnel or to the public in general.

5. Prudent Avoidance - Transmission

263 The following sections describe the detailed options for prudent avoidance mentioned in principle in Section 4 which may be applied to transmission facilities.

5.1 Transmission Lines

5.1.1 Distance

The most common method of reducing peoples’ exposure to EMFs is by selecting line routes (i.e. siting) to avoid population centres or areas where people gather. Particular attention should be paid to schools, child care centres and other areas where children congregate. Increased separation also needs to be considered when new residential development is proposed adjacent to existing transmission lines. Cost is a component of prudence, and in considering the feasibility of alternative routes or sacrificing land with significant development potential, regard should be had to the additional cost which may be involved.

Raising the height of supporting structures or towers, and thus the height of the conductors, can also reduce the magnetic field strength below the line and may also decrease the fields away from the line. The cost associated with the increased structure height may limit this technique to selected portions of a line. The approach may be more practical for wood pole lines than for steel tower lines, due to the cost factor. In some cases, raising height can increase the field strength at some distance from the line, and, accordingly it is necessary to treat each case on its merits.

5.1.2 Conductor Configuration

Different arrangements of phasing can produce different magnetic field strengths for the same line current. In general, triangular arrangements tend to provide more field cancellation than horizontal arrangements, with lower resultant field strengths.

5.1.3 Phase Arrangement

For double circuit lines, it is possible to arrange each three phase circuit with a different vertical phase arrangement in space, such that some cancellation of magnetic fields occurs. This is usually a relatively low cost option in the case of an existing line, and often a no cost option for a new line.

264 Selection of the proper phasing arrangement is usually the most effective way to reduce magnetic fields for two circuits on the same structure or two or more circuits on the same easement for minimal cost, if re-routing is not possible.

5.1.4 Split Phasing

A single circuit line can be constructed as two parallel circuits with a phase arrangement designed to achieve maximum field cancellation. This is known as the split-phase technique and may be considered if only one circuit exists on a route. Although this form of construction is significantly more expensive than conventional single-circuit construction, it could be used for short sections of a line where it is desirable to reduce fields within a cost limitation.

5.1.5 Current Reduction

A reduction in electrical current will generally reduce magnetic field strengths. The reduction in field strength is approximately proportional to the reduction in current. Transmission line currents (and voltages) are usually based on system stability and required load transfer considerations. The only practical way to achieve this option is to move to split phasing for a single circuit line, or a second line in the case of a double circuit line.

5.1.6 Shielding and Cancellation Loops

Shielding is the erection of a barrier between source and subject to reduce the field strength at the subject. Shielding can substantially reduce electric fields from transmission lines but has little effect on magnetic fields. Any object between the source (line) and the point of interest will provide shielding or distortion of the electric field. Common examples are buildings, trees or any other structure.

For all practical purposes there are no means to significantly reduce or screen magnetic fields from overhead lines. In special applications, screening of individual pieces of equipment is possible using structures or enclosures made from special metals, but these are expensive and limited in application.

5.1.7 Undergrounding

Because undergrounding is usually far more expensive than overhead construction, it does not often fall into the category of prudent avoidance, with its

265 “modest cost/minimum inconvenience” criteria. There will be occasions, however, when partial undergrounding may be consistent with prudent avoidance on a total cost basis, and therefore this option is considered here.

In underground cables, phase conductors are insulated from earth and from each other by a relatively thin layer of solid insulation as compared to a much larger dimension of air insulation in the case of overhead lines. Accordingly, underground phase conductors can be placed much closer together, providing a more effective field cancellation effect.

On the other hand, underground cables are normally buried one metre or less below ground and can be closer to people than an equivalent overhead line. Nevertheless, due to the cancellation effect, the use of underground cables usually reduces the effective level of the magnetic field at the point of interest. An exception to this might be the situation of cables in a street area where the point of interest is the footpath or roadway immediately above the buried cable where the field strength is still significant.

When considering undergrounding, it should be noted that, contrary to popular belief, earth has no magnetic field shielding property and plays no part in further field reduction.

A three phase underground cable in one sheath will produce a lower magnetic field than the same capacity line constructed from three single-core cables because the conductors are closer together and provide more effective field cancellation than three single-core cables, especially if they are in flat formation.

A very effective method of reducing magnetic fields from cables can be applied to a three phase cable constructed from three single core cables. This method of shielding utilises the principle of longitudinal electromagnetic induction where a longitudinal voltage is induced in a non-energised conductor that is in parallel with the current carrying conductor.

If each insulated phase of a three phase cable is laid inside a conductive conduit such as copper, aluminium or steel pipe and the pipes are electrically bonded together at each end, then each pipe will carry an induced current in the opposite direction to the phase current and hence provide some cancellation of the magnetic field. The effectiveness of this method is heavily influenced by the contact resistance of the bonds.

266 Practical experience shows that up to a 20 fold reduction of magnetic field can be expected from this measure. When applying this method of shielding it is important to derate the cable due to its reduced heat dissipation. This method is usually a high cost alternative due to the cost of encasing the cables in conductive conduit, and also because of the derating effect.

5.2 Transmission Substations

The magnetic fields produced by transmission substations result largely from the outgoing and incoming overhead transmission lines, especially where they come together at entries to busbar arrangements. Consequently, most of the prudent avoidance options available for transmission substations are those detailed above. Particular opportunities may arise where multiple circuits parallel one another as they converge on their substation. Also, there is sometimes an opportunity in the siting of the substation, or even in the fencing of the substation equipment (often part of the whole site) to keep minimised magnetic field strengths in areas accessible to the public.

5.3 Land Development

Land development adjacent to transmission lines often occurs after the transmission line has been built. In such cases, the application of prudent avoidance principles should become the responsibility of the developer. A developer should be expected to spend some money in reducing people’s exposure to magnetic fields, although detailed discussion of land development issues is beyond the scope of this paper. It is noted that the prospect of future land subdivision and development may create an argument for utilities adopting wider easements in the first place. In his 1991 Report, Sir Harry Gibbs found no support for such a move, which would alienate additional land and increase costs to the community. Gibbs said:

“.. it would be particularly undesirable at the present time to prescribe standards or guidelines with regard to exposure to the fields created by transmission lines or the width of easements acquired or used for such lines.”

All transmission line easements would be affected while any potential benefit would be restricted to a few isolated developments.

267 6. Prudent Avoidance - Distribution

6.1 Distribution Lines

6.1.1 Siting

Due to the need to provide electricity supply to customers, the options available to designers in siting distribution infrastructure are limited. Distribution lines by their very nature and function are normally located in road reserves to provide supply to customers on both sides of the road, although some instances, they are located at the rear boundary of residential properties.

Where practicable:

• distribution lines should be located on the opposite side of the road from areas such as:

• schools

• kindergartens

• child-care centres

• distribution lines should be sited away from the walls of multi-storey buildings or areas where children congregate

• distribution lines should be located on the side of the road bordered by open spaces where applicable

• substations should be located at the electrical centre of their low voltage network, i.e. current flows in all directions should be balanced.

6.1.2 Design

Options which may be considered, subject to their economic viability, could include:

• use of aerial bundled conductor (ABC) for low voltage reticulation to provide more effective field cancellation

268 • use of offset construction to increase horizontal separation from point of interest

• use of underground cable in place of overhead conductors where economically justified

• use of 3 phase cable instead of 3 single phase cables

• balancing of load across all phases to reduce neutral currents

• use of insulated twisted service cable instead of open wire services to provide more effective field cancellation

• for new double circuit lines, adoption of RWB/BWR phasing when current flow in both circuits is in the same direction

When installing electrical facilities which involve both low voltage and high voltage, the following options apply:

• When overbuilding (or underbuilding) existing facilities, the phasing on the existing circuits should be determined and the new circuit or circuits phased to minimise the combined magnetic field strength.

• Where new or reworked subtransmission facilities are being considered on the same structure with distribution circuits, the most effective field reduction measures may be applied to the distribution circuits.

6.2 Distribution Substations

6.2.1 General Principles

Distribution and consumer substations are typically 22,000/415 V or 11,000 /415 V and can generally be classified as:

• free standing open type ground mounted

• free standing enclosed type ground mounted

269 • underground (pit)

• indoor

• pole mounted

The main sources of magnetic fields from substations are:

• incoming and outgoing lines and cables

• busbars

• switchgear (ie. isolators, switches, circuit breakers, fuses, etc)

• transformers

• earth connections/neutrals where these can form alternative earth paths

Underground metallic pipes and telecommunication cables with metallic screens, or even structural steel can also be significant sources of magnetic fields if they constitute a return path for a portion of the substation earth or neutral currents.

The compact design of gas insulated switchgear (GIS), as compared to open or enclosed air insulated switchgear substations, offers significantly lower magnetic fields due to a substantial reduction of the phase separation distances. A degree of magnetic shielding is also afforded by the gas filled enclosures. The LV side of the substation has higher associated levels of magnetic field due to the higher currents compared to the HV side.

Metal-clad substations, where mild steel is usually used for fabrication of enclosures, are afforded a modest level of shielding by the enclosure. Also, reinforced concrete slabs, walls and floor panels can provide some magnetic field shielding. It should be noted that unless cables, busbars and the like are fully surrounded, any shielding afforded by metallic enclosures becomes less effective with increased distance from the source. Building materials such as brick, stone, plaster, wallboards and wood have no shielding properties for magnetic fields.

The following basic magnetic field management techniques can be applied in the design of substations:

270 • increasing the distance from source of magnetic field

• reducing the conductor or busbar spacing

• selecting an appropriate phase configuration

• balancing load between phases to reduce the neutral current

Magnetic field levels in excess of about 12 mG may cause interference to conventional computer monitors and the following recommendations for prudent avoidance can also serve to mitigate this effect.

6.2.2 Specific Measures

In designing distribution substations in situations where prudent avoidance is required, the following design measures may be considered. Some measures are more appropriate for high rise office situations, and some for outdoor substations near domestic dwellings.

In the case of high rise office buildings:

• locating substations away from normally occupied areas such as offices, lunchrooms, etc;

• planning the substation layout so that the LV side is further away from adjacent offices, computer rooms, etc than the HV side;

• locating transformers, LV busbars, disconnector switches and other potentially large sources of magnetic field within the area of the substation as far away as possible from adjacent offices, etc;

• if the floor above the substation is used as office space, avoiding where possible, direct ceiling mounting of heavy current cables, open type busbars or disconnector switches;

• locating all cable trays as far as possible from the substation ceiling and walls that separate it from adjacent offices, etc;

271 • avoiding the laying of heavy current cable directly on the floor of the substation if the floor below the substation is used for office space;

• designing busbars to minimise separation between phases and between phases and the neutral bus;

• if practicable, orienting transformers and other sources that have uneven field patterns so that their highest field strength side is turned away from the field sensitive area;

• where possible using three phase cables in preference to three single phase cables;

• using a trefoil arrangement of cables when using three single core cables in a three phase configuration. In such cases, if the neutral conductor is a separate single core cable, placing it, where practicable, in the centre of the trefoil formation of phases;

• selecting the substation equipment considering, among other important electrical parameters, its low magnetic field design, ie 11,000/415 V distribution transformers in steel housings, compact metal-clad busbars;

• avoiding phase by phase grouping of single core cables in parallel circuits;

• distributing all large single phase loads and all constant current load such as lighting and office equipment equally between three phases of the LV supply.

In the case of outdoor substations:

• positioning the secondary side of the transformer so that barriers such as landscaping, fencing or block walls inhibit normal access to that side of the substation;

• locating substations away from normally occupied areas such as bedrooms, offices, playgrounds etc;

272 7. Miscellaneous

Whilst the primary focus of this paper is on electricity supply business installations, sources within customers’ installations can also make a significant contribution to the overall magnetic field. Accordingly, a brief selection of considerations relevant to customer installations are provided in the following sections. Supply conditions may vary from utility to utility and, if inconsistencies are evident, these conditions should take precedence.

7.1 Commercial/Industrial Switchboards

In the case of large commercial/industrial switchboards, the busbars inside the switchboard can have an effect on field levels outside the switchboard. The following prudent avoidance measures may be available:

• Keeping the incoming line and associated meter panel and/or busbars away from heavy use areas. This will also help avoid computer interference problems.

• Avoiding the use of separate conductor trays for the energised and neutral wires. If separate trays are necessary, it is best to place them adjacent to low/no use areas.

• Locating switchboards away from high use office areas if possible

• Locating workstations away from switchboards when laying out new or reorganised office areas. A distance of 4 to 5 metres is suggested to provide the additional benefit of avoiding computer VDU interference.

• Use energy efficient lift motors, air conditioning equipment and industrial motors and manufacturing equipment.

7.2 Domestic Meters and Wiring

Generally, the principal source of magnetic fields associated with domestic meter boxes is the wires leading to the meter box. Accordingly, prudent avoidance measures associated with meter boxes generally focus on the wiring rather than on the box itself. The following options may be available:

273 • In general, for new constructions, the layout of meters, switchboard and wiring may be planned in advance, giving consideration to the magnetic fields that they would produce.

• Locating the meter box in an area that is not adjacent to high use areas. Good locations would be at the garage, a closet, storage room or at the back of a wardrobe. Bedroom and living room walls are better avoided to reduce fields in active use areas. Many authorities recommend the placement of meters and switchboard in a back-to-back arrangement, with meters outside and switchboard inside the home for security of home and occupants. This arrangement usually places the switchboard in low-use areas (for the sake of appearance), and is consistent with prudent avoidance.

• Locating the main connecting wiring away from high use areas in cases where meter location and switchboard location are separated by a significant distance, e.g. where meters are installed at the fence and the switchboard is located at (or in) the house. The connecting wiring should be run with phases and neutral grouped together, and in a ceiling space rather than a wall space, for example.

• Using service wires of insulated twisted construction, as they produce significantly weaker fields than open wire (bare conductor) construction.

• Minimising or avoiding situations where heavy current wiring, especially that of stoves and air-conditioning is placed in wall cavities within the house. This type of wiring is best located and grouped together in the ceiling. Close proximity of the phase wires and neutral helps to cancel the magnetic fields.

• In the case of two-way switches, running the neutral wire along the same path as the twin active wire connecting the two switches to provide a cancelling effect on the magnetic fields.

• Using energy efficient equipment which will use less electricity and save money, as well as reducing the electrical load on the switchboard, thereby reducing magnetic fields. Large white goods such as refrigerators, dishwashers, washing machines and driers are often sold with energy efficient model alternatives

274 7.3 Earth Connections

The Multiple Earthed Neutral (MEN) system is commonly used to connect a utility’s neutral at a customer’s switchboard. This neutral is earthed at the switchboard, sometimes via a metallic earth stake, and sometimes via a metallic water pipe. Depending on the condition of these earth connections, some fraction, or indeed, the majority of the neutral current may flow through a path other than via the utility’s neutral. If this happens, then an earth connection or water pipe can itself become a substantial source of magnetic field. The supply wiring also becomes a source of significant magnetic field as the magnitudes of the active and neutral currents are not equal.

When considering options for earth connections, it is the responsibility of the owner/electrician to identify and implement appropriate actions as follows:

• If metallic water pipes are a source of magnetic fields, install a plastic joint at the entrance to the building’s water system to prevent current from adjacent earths travelling along the pipe. In this case, the installation of a separate earth stake is mandatory for electrical safety considerations.

8. Conclusions

The concept of prudent avoidance has been recommended as the most appropriate public policy response to health concerns associated with magnetic fields. Because by its very nature it is not well defined in precise terms, historically it has been difficult to scope. Nevertheless, it is possible to adopt many specific measures which are consistent with the notion of doing what can be done at modest cost and without undue inconvenience to reduce people’s exposure to magnetic fields. This paper has sought to discuss the concept of modest costs and to suggest a range of options or measures for transmission and distribution applications. This paper has also touched on a number of options which may be adopted by other stakeholders in this issue such as electricians and home and building owners.

ACKNOWLEDGMENTS

The authors wish to acknowledge the assistance given by many colleagues within the Australian electricity supply industry in the development of this paper.

275 REFERENCES

Gibbs, H. “Inquiry into Community Needs and High Voltage Transmission Line Development”, NSW Govt. (1991).

Melik, Garry. “Magnetic Field Mitigation to Reduce VDU Interference”. Electricity Supply Association of Australia Limited. (July 1996).

Morgan, Granger M. “Electric and Magnetic Fields from 60 Hertz Electric Power: what do we know about possible health risks?”, Department of Engineering and Public Policy, Carnegie Mellon University, Pittsburg, PA (1989).

Morgan, Granger M. “Power-Frequency Electric and Magnetic Fields: Issues of Risk Management and Risk Communication”, In: Biological Effects of Electric and Magnetic Fields: Beneficial and Harmful Effects, edited by D. O. Carpenter and S. Ayrapetyan. Volume 2. Academic Press (1994).

The Panel on Electromagnetic Fields and Health. “Report to the Victorian Government”, Vict. Govt. (1992).

Senate Economics References Committee. “EASTLINK. The Interconnection of NSW and Queensland Electricity Grids with a High Voltage Powerline”, Australian Govt. (1995).

Texas Public Utility Commission. “Health Effects of Exposure to Power-Frequency Electric and Magnetic Fields”, (1992).

ADDENDUM

Policy Statement on Electric and Magnetic Fields (EMFs) (Approved by the ESAA Board on 27 March 1997)

Background:

The Electricity Supply Association of Australia Limited (ESAA) is the national voice for members of the electricity supply business in Australia. ESAA is committed to taking a leadership role on relevant environmental issues including power frequency electric and magnetic fields (EMFs). ESAA and its members are

276 committed to the health and safety of the community including their own employees.

Adverse health effects from EMFs have not been established, but there remains a lack of scientific consensus about whether or not they can occur. Because of this lack of scientific consensus, the question of whether EMFs can cause adverse health effects such as cancer is important both for the Australian community and the electricity supply business. ESAA recognises that there is concern within the Australian community about EMFs and is committed to addressing community concerns by the implementation of appropriate policies and practices.

ESAA is committed to a responsible resolution of the issue where government, the community and the electricity supply business have reached public policy consensus consistent with science.

Policy Statement:

1. ESAA recommends to its members that, within Australian health guidelines, they design and operate their electricity generation transmission and distribution systems prudently1.

2. ESAA will closely monitor engineering and scientific research including reviews by scientific panels, and overseas policy development.

3. ESAA will communicate openly with all stakeholders including assisting its members in conducting community and employee education programs, distributing information material including newsletters brochures booklets videos and the like, liaising with the media and responding to enquiries from members of the public.

4. ESAA will co-operate fully with any bodies established by governments in Australia to investigate and report about power frequency electric and magnetic fields.

277

The Politics of EMFs - An Activists Viewpoint

Richard W. Woodley Webmaster, Bridlewood Electromagnetic Fields (EMFs) Information Service Ottawa, Ontario, CANADA

Remarks to Public Viewpoint Panel

What is the public viewpoint on electromagnetic fields? From my perspective as an activist I see it as one of confusion. Confusion by reading about study after study pointing to a link between EMFs and health risks such as cancer and leukaemia and seeing contradictory statements from various agencies saying “there is no definitive proof” of such risks.

Why is the public subjected to this? It is because of the politics of EMFs. The politics of EMFs that sees reports linking EMFs and health suppressed by the White House. The politics of EMFs that sees agencies releasing reports with press releases that downplay or even contradict the findings of health links that are contained in the reports.

The latest piece of confusing information the public has been exposed to is the National Institute of Environmental Health Sciences panel finding that EMFs are a “possible” human carcinogen. What does that say to the public? Nothing! Bombarded with news stories about studies linking all sorts of things, including common foods, to cancer the typical public response is “So what - everything causes cancer these days.”

However, if the truth be known, in my humble opinion, that is not the message the panel intended to convey. I believe they would not have examined the massive amount of reports and studies out there, and agonized over them just to say something as innocuous as that. I believe that they really want to convey the message that EMFs are PROBABLY, not possibly, a human carcinogen, but the politics of EMFs prevented them from going that far.

What the public deserves to hear is the truth. Not everyone exposed to EMFs is going to get cancer, indeed most won’t. But there is now overwhelming evidence that those who are exposed have a much greater risk of cancer, leukaemia and other health problems. Indeed the studies indicate that children, in particular,

279 exposed to higher EMF levels, have two to three times the likelihood of getting cancer or leukaemia.

What the public wants to know is - What are we going to do about it? When are we going to do the research necessary to find out how EMFs are linked to health problems so we can find out what we need to do to reduce the risks? And when are we, as a society, going to start taking serious prudent avoidance measures?

We know we cannot eliminate the risk, just as we cannot eliminate the risks associated with driving automobiles, because no more are we going to ban electricity than we are going to ban cars. But we can take measures to reduce it, especially when children are concerned. For example, nobody would allow a 100 km/h highway beside a school. So why do we allow high voltage power lines beside schools?

So what is holding us back? It is the politics of EMFs. The public deserves better than to be subjected to the politics of EMFs.

280 Rapporteur’s Report

Tom McManus, Ph.D. Chief Technical Adviser, Department of Public Enterprise 44 Kildare Street, Dublin 2, Ireland Tel: +353 1 604 1023, Fax: +353 1 604 1189 E.mail: [email protected]

The seminar was opened by Dr. Elizabeth Nielsen of the Radiation Protection Bureau of Health Canada. She mentioned the wide range of interests represented among the participants and thanked each of the nine sponsoring organisations whose generosity had made the conference possible.

The chairman, Dr. Michael Repacholi of WHO, reminded the participants of the purpose of the meeting. As a key part of its International EMF project, WHO planned to publish a monograph on how the concepts of risk perception, risk communication and risk management can be applied to the EMF issue. The output of the two-day scientific programme was to discuss the main aspects of these issues and to contribute to the working group meetings which were to follow the seminar.

The keynote address was given by Dr. John Graham of the Harvard School of Public health. Professor Graham's talk was the kind you once gave up your lunch break for when you were a student. Although the title of his talk was "Making Sense of Risk", it soon became clear that the public and the media had a poor understanding of risk and especially in judging relative risks. The public judge risks by the ease with which it can imagine them while the media are driven by rarity, novelty and drama.

Dr. Graham discussed two kinds of bias. The first is the optimism bias where, for example, 80% of us believe we're better than average drivers. Then there is the outrage bias where in matters that are beyond our control, like nuclear power and chemical pesticides for example, the policy embraced is one of 'better safe than sorry'. Professor Graham felt that EMF was moving into outrage territory.

The public's lack of intuition for numbers is exposed in how it underestimates the likelihood of accidents that can arise from disjunctive events while it overestimates the risks from conjunctive events. Disjunctive events are those like car accidents where the risk due to one particular cause, like brake failure, is very small but the number of possible causes of accidents can be measured in thousands and so the overall risk of accident can be significant. Conjunctive events, on the other hand, are those where the accident or event only happens when a sequence of events, where each one of which might carry a measurable risk, all must all occur simultaneously or in a particular order.

Public perception, or should it be misperception of risk is illustrated by the fact that while Americans are healthier and living longer than ever some 75% believe life is riskier today than it was 30 years ago.

Attitudes to risk of a sample of 1019 lay people were compared to those of 264 members of the Institute for Medicine. Using the HCRA Confidence Scale it was found that while 40% of the lay group considered EMF a concern (where the HCRA score was 7 or more) only 15% of the medical group shared this view.

A profile of an EMF believer could be drawn:

• female, egalitarian, with children, living in the northeast United States.

The sceptic's profile was:

• well-off, highly educated, male, who understands actuarial mathematics.

So what can be done to change these misperceptions? Dr Graham recommended :

• Improved risk assessments; existing ones are not always reliable.

• The need to better explain the hazard and the magnitude of the risk .

• The need to rank risks in order of priority.

• That we remind the public that what reduces one risk can create a new one, (e.g. passenger side airbags in motor cars)

His final message was a plea for more scientists and engineers doing risk assessments for government agencies.

282 Following Professor Graham, Dr. Repacholi reviewed the objectives of the WHO's EMF Project and the framework within which its various outputs were being produced. He discussed the work of the International Advisory Committee which brings together all national authorities involved in the EMF - health issue. He discussed WHO's collaboration with IARC on matters such as electrical hyper- sensitivity and gave a timetable for the production by WHO and IARC of their respective assessments of the health effects of ELF and RF fields.

Dr. Repacholi then reviewed the basis for the ICNIRP guidelines for public and occupational exposure to ELF and RF fields and concluded by discussing the implications of the NIEHS Working Group's report on power - line frequency fields published in July 1998.

Dr. Philip Gray was the first of two speakers from the Federal Research Centre at Juelich in Germany. Dr. Gray discussed the need for analysis and deliberation in improving risk management and risk communication. It was this paper that brought it home to the participants of how simple things used to be when there was only the triumvirate of science, industry and government to consider. Science came up with ideas, industry turned them into money, and government made the rules and collected the taxes. Now people - the public - have got in on the act and must be considered. So with people now involved in most decision-making processes Dr. Gray provided a host of good advice on how this new player should be handled:

• People should have input into risk decisions that affect them.

• Their risk perceptions can enrich expert analysis.

• Their perceptions reflect basic values.

The social dimension in dealing with the public is of prime importance. There is the importance of trust in mediating and interpreting scientific information for the public. The public were now more likely to promote what once might have been thought of as just some simple danger into a real actual risk.

Two risk management frameworks were discussed - the analytic- deliberative framework of NRC and the risk management cycle produced by the US Congress Committee on Risk Management. The latter framework lays greater

283 emphasis on stakeholder collaboration and the US RAPID Program was seen as a good example of this.

Following Dr. Gray's paper, his colleague at the Federal Research Centre, Dr. Peter Wiedemann, continued the theme in a discussion of how one might deal with EMF risk perceptions. He began by reiterating many of the awkward questions the public ask about EMF and he discussed strategies that might be put in place to answer them. The problem is that while experts evaluate information by scientific methods, lay people use intuition. Dr. Wiedemann noted that the most difficult individuals to deal with are the electro-sensitive. He concluded his paper with a very useful and detailed guide on preparing for a public meeting on the EMF issue. Preparation and planning could not be over-emphasised. Some suggested techniques for handling and diverting 'loaded' questions were particularly well- received.

Mr Gerry Kruk of Gerry Kruk & Associates communications of Calgary was the next speaker. He reminded the audience of the good old bad old days.

In Calgary it was an energy company and the project was a sour gas well development. The company made little progress when at first it tried to browbeat the Energy Resources Conservation Board. Mr. Kruk's talk illustrated that you can't really go too far wrong when you treat the public as human beings living on the same planet. He took us through the various steps taken by the energy company once it had seen the light and changed its approach. The changing nature of the newspaper headlines as the campaign proceeded made fascinating reading.

Ms. Judy Larkin of Regester Larkin (Consultants), London provided an excellent insight into public attitudes in the UK concerning the health risks of mobile phone use and what the phone companies are doing to counteract this threat to their business. As a result of the many health scares in the UK over such matters as BSE, e-coli, HRT, breast implants and Brent Spar, the public's regard for and trust in industry and government was at an all time low. In contrast activist groups were now dominating television coverage of the cellphone health issue. Even in the newspapers activists were getting 50% more coverage than industry spokesmen. Scientists working for environmental groups are now more trusted by the public than either government or industry scientists.

Ms. Larkin's conclusion was that industry must be proactive and more actively involved as an expert. Unless this happens there was a danger that the UK

284 government might be forced into some action over cellphones, not in response to scientific findings but to a growing public anxiety. She concluded her talk by quoting Woody Allen: "The world is run by the people who show up".

Professor William Leiss of the School of Policy Studies at Queens University, Kingston, introduced his presentation with the largest number of boxes ever seen on a slide. It turned out to be a guide to the Internet around EMF and particularly the RF sub-category. Professor Leiss' paper came across as a warning to all those in the 'establishment' that the public were out there and that they could get all the information they needed within hours from the Internet. No longer had they to depend on what governments or Health Canada decided to tell them. Armed with this, information activist groups were now able to make life extremely difficult for those wanting to put up new phone towers, or hide old ones inside church steeples in Vancouver.

And as far as public outrage was concerned "We ain't seen nothing yet", was Dr. Leiss' view. Next year local multipoint broadband telecommunications systems are due to start operations in Canada. These loop systems will require even more antennae towers. Dr. Leiss' message was "Be warned. Stop denying there is a risk. Stop claiming there are no biological effects. Because there are!"

The final paper of the first day was by M. Andre Beauchamp of Enviro-Sage in Montreal. He reviewed the history of Hydro-Quebec's dealings with the Quebec Dept. of Health, the Bureau d'Audience Publique sur l'Environment (BAPE), and the Quebec government over its various power line projects since 1983.

BAPE is an interesting idea for Anglo Saxon regulatory authorities, in that it provides a forum for public debate in advance of a decision by the authorities on a project. It seems however that the public doesn't get too much involved and that scientists from the Department of Health are the main source of critical contributions to BAPE on EMF issues.

It is interesting that following last winters' ice-storm Hydro-Quebec is proposing to build three new supply lines into Montreal and hopes to have construction underway before this coming winter. Because of the critical need for these lines the Quebec government has suspended the need for an environmental impact assessment and is adopting a streamlined approvals procedure. There are critics of this approach. M. Beauchamp sees the attitude to EMF risks in Quebec

285 changing. Where once the risks were considered merely plausible, they are now regarded as uncertain, if not probable.

The second day's session commenced with an excellent paper by Dr. Bill Bailey of Bailey Research Associates. He invoked native American medicine men, Polynesian shamans, and his pet retriever to support the need to put risk communication at the start of any risk assessment process and not leave it as afterthought to be added like credits to the end of a movie.

In listing the qualities that make good risk assessment support good risk communication Dr. Bailey mentioned the importance of thorough documentation and a transparent evaluation of scientific data.

He also scored the INCIRP guidelines and the NIEHS report against his criteria for risk assessment. The probabilistic approach to establishing the likely risk of a specific adverse health effect has much to commend it and could be looked into by standards setting authorities. There appears to be unnecessarily large margins of safety in the current standards.

Dr. Bailey was followed by Professor Daniel Wartenberg of the Robert Wood Johnson Medical School, New Jersey. He provided what was essentially a one-man seminar on the scientific approach to inferring risk. Weight of evidence, meta analysis and quantitative risk assessment are the three pillars on which this approach is based. It is easy to see why weight of evidence is used in the law courts. Meta-analysis appears to be a minefield. Meta-analysis should be able to survive the removal of any one particular study from the analysis.

He presented his own risk assessment of the NIEHS working group report data. From this Dr. Wartenberg concluded that with a risk ratio of around 1.3 about 6% or 100 to 200 childhood leukaemia cases may be associated with power lines.

Professor Wartenberg was followed by Dr. Chris Portier of NIEHS. Dr. Portier provided the inside story to the background and development of the NIEHS Working Group Report. The various votes of the 30 - person scientific committee and how these squared up against the IARC criteria gave a fascinating insight into the workings of the group.

The conclusion of the working group - that EMF is a possible carcinogen - has already received world-wide publicity. However it is important to note that the

286 working group report is only one of seven separate inputs which Dr Portier must take into consideration when he and his colleagues prepare their final report to Congress. Dr Portier concluded his presentation by arguing that the findings of the National Academy of Science and the NIEHS Working Group are not in such disagreement as would appear at first sight.

The morning session was brought to an interesting close with a talk by Dr. Caron Chess of Rutgers University. She provided a stimulating assessment of the pros and cons of two of the 9 forms of public participation she listed - public meetings and Citizen's Advisory Committees. Many who dislike or fear public meetings find favour in the Citizen's Advisory Committee idea. The bottom line from Dr. Chess's paper is that: "To be effective, public participation must be public participation based on research and not on accumulated wisdom".

In the final session of the seminar Dr. Tom McManus from Ireland was the penultimate speaker. He outlined how the EMF issue is handled in Ireland. In some areas - powerlines - some progress is being made. In other areas - mobile phone towers and dealing with the electrosensitive - lots of things have been done but the problems are growing, not diminishing.

The final presentation was delivered by Mr. Michael Dolan of the Electricity Supply Association of Australia. Mr. Dolan talked about prudent avoidance - a policy that Australians have warmly embraced. In 1991 the policy was proposed for New South Wales. It was then endorsed by the government of Victoria in 1993 and then by the Australian Federal Government in 1996. In its application, prudent avoidance measures continue to follow the guidelines set out by Sir Harry Gibbs in his 1991 report to the NSW government.

In its seven years experience of applying prudent avoidance in Australia and measured in the context of public acceptance ESAA believes it has proved to be a worthwhile common sense approach and the best policy option available.

287 288 289 290 291 WHO/SDE/OEH 99.01

292