Kristin S. Shrader-Frechette, Risk and Rationality (U. California Press 1991) [312 pp.] Index of Names, Index of Subjects, Notes. CIP 91-3294; ISBN 0-520-07287-1 (Cloth $39.95); ISBN 0-520-07289-8 (Pbk. $15.95). [2120 Berkeley Way, Berkeley, CA 94720.]
The problem presented by large scale activities that expose unknown people (the "public") to risks of uncertain range, duration, and severity is clear: A market mechanism is not even arguably able to solve such problems, and a second best solution will inevitably result in decisions by one set of people that expose others unwittingly and unwillingly to risk. Professor Shrader-Frechette argues that the two existing candidates for second best -- naive positivism and cultural relativism -- are critically misguided. Where naive positivism attempts value-free objectivity, it succeeds only in burying value choices in the complexity of its analysis. And cultural relativism, which rejects the possibility of objective analysis where values are concerned, condemns the entire activity to simple choice, unilluminated by rational discourse.
Professor Shrader-Frechette's proposal, "scientific proceduralism," is based upon the proposition, which she strongly argues, that rationality is possible in value-laden decisions. Risk-cost-benefit analysis (RCBA) might, for example, form the basis for risk evaluation, but, where that is the end of the process of evaluation to the naive positivists, she would subject it to adversarial assessment from alternative ethical systems. A given RCBA might, then, be evaluated and argued from the standpoint of Benthamite utilitarianism, Rawlsian egalitarianism, Nietzshean elitism, or Paretian libertarianism. Each ethical system would provide a ground from which the RCBA could be assessed. The procedure would be competitive, each advocate arguing the strengths of one interpretation over the others.
Professor Shrader-Frechette devotes the last two chapters of the book to suggestions for implementing scientific proceduralism. The idea has obvious applicability in the regulatory process, where RCBA is ubiquitous. Its implementation in the political arena, which is significant because of the decentralized nature of American land use decisions, is less clear. It is in her discussion of the judicial process that her ideas are to me, perhaps because I am a lawyer, most compelling.
On the face of it the judicial process is not a likely prospect for the implementation of scientific proceduralism. Only when it passes on an appeal from a regulatory or political decision is the judicial process involved in a risk decision prior to the creation of the risk, and then its review is narrowly procedural. Courts are generally involved after a risk has manifest itself in injury, as in the case of environmental tort litigation. That is hardly the time or place to implement scientific proceduralism, though one might imagine that if it were required, rather than RCBA, the failure of the defendant to implement it properly might be the basis for a finding of fault.
The judicial process comes into this story in a different way, according to Professor Shrader-Frechette. Were those injured by an environmental risk better able to prove the case, the costs of an environmentally risky project would be forced onto the actor, who would then have to internalize them. Knowing that it would be forced to internalize the cost of injuries, the actor would be forced before undertaking a risky project to put its money where its mouth was: To undertake only those actions for which it could afford to adjust the cost of its product or service. Any inaccuracy in its RCBA would then come back to haunt it, for it would either fail to undertake projects because of overstated RCBAs, or undertake projects that turned out to be more costly than the RCBAs had predicted. Under such circumstances, one might expect the people involved to act responsibly.
But how might the judicial process be made efficient enough to make such discipline on risky actors effective? Professor Shrader-Frechette mentions a variety of proposals. The fundamental problem is that conventional tort law is not the ideal place for claims by those who have been injured by environmental risks. Tort law looks for fault on the part of the actors, and in these actions the actors are liable to be able to show extraordinary expense and effort bent to assessing and avoiding harms. What is needed is a cause of action in which the plaintiff need only prove that a recognized risk caused harm. The plaintiff's recovery would be proportioned to the statistical contribution that the defendant's actions made to the likelihood or severity of the injury. Such a procedure would force the defendant to internalize all of the costs of its actions, not simply those that resulted from fault.
In the hands of an entrepreneurial tort lawyer Risk and Rationality could be a most potent force for the emergence of a new flavor of tort law. Were such a theory to emerge, however, it would reduce the need for scientific proceduralism for it would approximate a Pareto-like solution to the problem of risk. An action is optimal under the Pareto criterion if everyone is either unaffected by it or made better off. Were firms to be forced by the law to compensate everyone who lost from a risk to indifference, it is at least arguable that they would undertake only projects that satisfied this criterion.
It is not likely, however, that Professor Shrader-Frechette would find this judicial solution satisfactory. One reason is that it would not, in fact, result in a Pareto optimal solution, for it would force costs upon people, then pay those costs. But is suspected that her objection would go deeper. There is in her analysis the rumblings of two values that would not be satisfied by this solution. One is her suspicion of the willingness of those affected by decisions. Throughout this book, she is concerned that what appears to be willing acceptance of risk is either somehow biologically predetermined (i.e., risk preference) or coerced by circumstance (i.e., poverty). The second weakness that she might find is that it avoids citizen participation. She quotes Thomas Jefferson to the effect that an ignorant citizenry should not be denied the power to make decisions; rather, they should be informed. There is, one suspects, a virtue in public decision-making beyond the plain virtue of better decisions about risk.
Risk and Rationality is not an easy read. Professor Shrader-Frechette makes a succession of powerful points and resolves them with an important proposed change in the way decisions are made. Her argument is throughout spare and clean, her prose free from didactic asides, and the whole work is illuminated with pithy descriptions of actual controversies. It is demanding but leaves the reader grateful that such demands were made.
Hugh H. Gibbons* Mr. Gibbons is Professor of Law at Franklin Pierce Law Center
Top of page Index to Risk Book Reviews
W. Kip Viscusi, Reforming Products Liability (Harvard U. Press 1991) [270 pp.] Appendices, bibliography, index, notes. CIP 90-23161, ISBN 0-674-75323-2 (Cloth $39.95). [79 Garden Street Cambridge, MA 02138.]
It is clear that Professor Viscusi recognizes, even insists upon, his perspective: "I view products liability reform... as a question of designing an appropriate institutional mechanism to control risks and compensate victims of product-related injuries." But he does not assess products liability law as if that were its purpose; to Professor Viscusi, that is its purpose. For all its empirical tone, the central message of this book is a normative one: If it is not, in fact, the purpose of products liability law to control risks and compensate victims -- and there is much of products liability law that is nonsensical from that point of view -- it should be.
This book is the coming home to roost of one of the most problematic propositions in Richard Posner's seminal Economic Analysis of Law. From the standpoint of safety, according to Posner, it did not matter whether the law applied a negligence standard or an absolute liability standard to products. If it applied a negligence standard, producers would take cost-justified loss-avoidance. The cost of accidents would rest upon those who were injured by irreducible risks, for they could not show that the manufacturer was negligent. Under an absolute liability standard, firms would take the same level of cost avoidance, though they would have to pay those who were nonetheless injured. There would be a distributional difference between the doctrines (absolute liability would make victims, and their lawyers, richer), but no allocative difference between them; the injury rate would be the same under both regimes.
So what? If the purpose of products liability law was to optimize the injury rate, we would be indifferent as between the two doctrines. If the purpose were to compensate victims, law would favor absolute liability. It follows, then, that if the purpose of the law was both optimizing the injury rate and compensating victims (which are the two purposes that Dr. Viscusi finds underlying products liability), there is an ideal rule: absolute liability. The law, they would predict, will move toward an absolute liability rule as soon as it gets its act together.
Courts have indeed toyed with the idea of absolute liability (and its vacuous kin, "strict liability") over the past thirty years but have resoundingly rejected it in every state. That might suggest to economists that there is a dominant purpose lurking in products liability law which has thus far escaped them. I suspect that this has, indeed, dawned on Professor Posner, now a judge on a federal court of appeals. But it seems to have altogether escaped Professor Viscusi, for he adamantly grounds his analysis on controlling risk and compensating victims.
What then is the deeper purpose that could account for the law's rejection of absolute liability? The answer to that question, unfortunately requires a different way of thinking than that employed by Professor Viscusi. His thinking is what is called "instrumental" or "consequentialist." That is, he views law as an instrument to be used toward the provision of a larger social objective. In this case, products liability should be an instrument for delivering "an appropriate institutional mechanism to control risks and compensate victims." It is no accident that the purposes toward which he imagines products liability law bent happen to be quantifiable and therefore amenable to the kind of quantitative analysis that his discipline prepares him for.
There is another way of thinking about law referred to as "formative reasoning." Here, the law is viewed as a process for forming the normative sense of the members of a society. By publicly recognizing and documenting wrongs, the law reinforces the sense of responsibility in its members and helps shape it to new realities. The sense of right and wrong, under this view, is an emergent phenomenon, as the law itself gropes now left, now right, attempting to unearth the responsibilities that humans have in new situations.
The history of products liability law is particularly supportive of such a view of law. Negligence law grew out of the context of direct human interaction, of people accidentally hitting each other with sticks, running over pedestrians, or docking their boats at piers they did not own. By the end of the nineteenth century, however, the dominant mode of economic organization had become the firm. Decisions made within the firm would affect people whom the decision-makers never knew and cause injuries to people whom they never saw. Worse, those who were injured had a devil of a time proving who had done them wrong. The tort law of the time was distinctly people-oriented, so the plaintiff had to bring the specific wrongdoer into court. With the advent of the corporation that was not easy.
The corporation itself had a formative effect upon people, but a distinctly negative one. By wrapping the decision-maker in layers of organizational protection, it removed the person from the immediate context within which his sense of right and wrong was triggered. It was entirely possible to feel very good about oneself, to treat wife and neighbors with exemplary kindness, while adopting corporate policies that would strike limbs from the workers and poison the environment.
It became necessary for law to penetrate such a context, to open it to scrutiny. Law has several mechanisms for doing this, and it used them all -- antitrust, rate regulation, safety regulation, and so on. Products liability proved to be a particularly handy mechanism, as Professor Posner has pointed out, because it made every person his or her own attorney general, penetrating the corporation in pursuit of redress of wrong. Such a task is not, however, an easy one. Products liability cases are among the most difficult to prove, for the defendant generally has many orders of magnitude more resources to use, and it controls the facts that plaintiffs need to prove. The law needed to expand the products liability plaintiff's power. How to do it?
In retrospect, the law had a mechanism that would have been ideal for the job, the doctrine of res ipsa loquitur. This doctrine shifts the burden of proof to the defendant to prove that it was not negligent upon the plaintiff's showing an injury that was probably caused by someone, somewhere committing a wrong. But in the seminal California case, Escola v. CocaCola, Judge Traynor, rejecting the direction of the rest of the court set on the path of applying res ipsa to products cases, opined that this was a good area for absolute liability. Where res ipsa would increase the plaintiff's power by reducing its burden of proof, absolute liability gave the plaintiff the incentive to file suit by promising a pot of gold, whether the defendant was at fault or not.
Absolute liability was a lousy idea, for it compromised the normative foundation of tort law by promising compensation without fault; it threatened to eviscerate the morally formative power of tort law, turning it into a weird form of insurance. As with all bad ideas in law, it eventually died, but not until it had excited plaintiff's lawyers to a frenzy of creativity and had shaken corporate executives to their roots. The confusion that it caused was probably worthwhile, for it got the attention of those most deeply buried in the corporate cocoon, threatening to expose their most private memos to the fresh breeze of discovery.
A formative perspective would predict that judges would refuse to yield the idea of fault, and that is precisely what has happened. The doctrine of absolute liability was explicitly rejected early on, and in its place came the rhetoric of "strict liability" -- from its inception, a name without a notion. There is no intermediate point between negligence liability, that requires the plaintiff to prove causation by fault, and absolute liability, that requires only proof of causation. The rhetoric of strict liability did, however, let judges capture the idea that, where products were concerned, they were not going to apply garden-variety negligence law. Today, it is safe to say that, in most jurisdictions, products liability is negligence law, but the burden shifts to the corporate defendant upon the showing of a "defect" without requiring the plaintiff to identify the individual responsible for the defect.
I suggest, then, that the formative purpose underlying products liability law fits the actual history of this law far better than Professor Viscusi's instrumental purposes (injury reduction and victim compensation). Now it is clear that the two types of purpose are related. If products liability induces a sense of responsibility in otherwise isolated corporate officials, one would expect that to show up in safer products. Safer products are, however, an artifact of the underlying purpose. Victim compensation, by contrast, is the engine that drives the responsibility-producing system. By offering the successful plaintiff the potential for a financial bonanza, the law creates an incentive for plaintiff's lawyers to penetrate to the deepest layers of corporate insulation. From an instrumental perspective, capping injury awards would save little of the cost of products liability, as Professor Viscusi points out. From a formative perspective, it would drastically reduce the incentive to prosecute a case that is hard to prove. It is a very bad idea, not because it would cause accident rates to rise but because it would create an incentive for corporations to hide their actions under ever deeper layers of protection, making cases against them ever more difficult to prove.
The difference between the instrumental and formative views is best illustrated by the Ford Pinto case. To Professor Viscusi the problem in that case was that Ford based its valuation of life on damage awards by courts, which were too low. As a result, it severely undervalued the losses that the relocation of the gas tank would avoid in its cost-benefit analysis. Had Ford calculated the losses correctly it would have found them to be ten times greater than the costs of making the change. Ford goofed.
That would explain why Ford would have to pay compensatory damages, but it doesn't explain colossal punitive damages, later reduced, assessed against Ford. Clearly the thing that burned up the jury was the very fact that Ford was calculating the value of human life. Doing so, it might be felt, is like a murderer who feels fine about killing a sick person because he knows that the value of the sick person's life is so low that he can afford to compensate for it. To the jury, what Ford did was evil.
It is at precisely this point that the formative and instrumental views clash, and I am tempted to side with the instrumentalists. Products liability law requires that each person act responsibly. A corporate officer is, however, responsible both to his or her customers and shareholders. If expenses are to be increased, those expenses must be justified, and justified in terms of corporate finance. Doing a cost-benefit analysis does not evince a coldness of spirit but rather responsible corporate behavior. Yet it does evince an impulse to treat human beings as objects, rather than subjects, so is suspect.
It must be said that Professor Viscusi's analysis, in the best traditions of economic analysis of law, sheds considerable light on the operation of the system. Particularly telling is his analysis of the insurance crisis (attributable, in substantial measure, to insurance practices rather than products liability law) and hazard warning law (for which he proposes a sensible general scheme). Odd, however, is his entertainment of one of the worst ideas to make itself onto the products liability scene, Dean Wade's risk-utility test. What, one might ask, is a court doing evaluating the utility of a product? Is that not the quintessential job for markets? Professor Viscusi does indeed ask that question, even shows that it is impossible to administer as a judicial test because it is mired in Kenneth Arrow's indeterminacy of individual values, but proceeds nonetheless to accept it as the most rational of the tests of products liability. One suspects that Professor Viscusi is resident on the utilitarian side of the economic analysis aisle.
Lest this seem to be faint praise for a serious work of scholarship, let me say that I found Professor Viscusi's empirical analysis illuminating, clearly grounded and helpful, particularly in pointing to what appears to be an emerging reinterpretation of law.
Hugh H. Gibbons*
* See previous review
Top of page Index to Risk Book Reviews
Biotechnology: Assessing Social Impacts and Policy Implications (David J. Webber ed., Greenwood Press 1990) [239 pp.] Bibliography, contributor data, index, notes, preface, tables. LC: 90-2935; ISBN: 0-313-27454-1. [Cloth. $42.95. P.O.B. 5007, Westport, CT 06881]
The first part of the book addresses the social, political and even theological dimensions of biotechnology, but mostly within the U.S. For example, the essay treating theological perspectives is limited to consideration of Judeo-Christian perspectives.
The second part of the book discusses responses of three types of institutions to biotechnical development: American universities, domestic agricultural cooperatives and international agriculture research centers. Until recently, most agricultural innovation could not be protected by patents and other forms of intellectual property, but papers in this part reveal that changes in the law have had a major impact by shifting R&D funding from the public to the private sector. For example, some university administrators have become interested in "technology transfer," and one essay presents an empirical study of factors influencing university biotechnical patent activity. Among its findings is that: A professional staff is useful if more patent activity is desired, but patent activity is inhibited if academic career advancement primarily depends on publication of peer-reviewed primary research.1
In the third part of the book, the effects of proprietary rights in biotechnology are also discussed, but the main concern is the effects of, e.g., the need for regulatory compliance. While the first essay is quite broad, the second and third focus on federal regulation and examine, respectively, the likely impact of regulation on the use of genetically engineered organisms in dairy fermentation and on the use of growth hormones in milk and meat production.
Biotechnology ends with discussions of the evolution of policy debates from whether to how biotechnology will be used and the responses of federal and state officials. The concluding chapter, written by the editor, challenges policy analysts to be more responsive to the needs of Congress and challenges Congress to be more alert to as well as to use "policy knowledge."
Unfortunately, the potential utility of this book is diminished in two ways. First, while Webber acknowledges that legal and life sciences perspectives also must be considered in addressing future uses of biotechnology, his book offers scant treatment of more direct impacts on humans -- ones that may prove to transcend those raised by agricultural applications.2 Further, while projected behavior of the private sector plays a major role in the book, papers were written from what appear to be primarily governmental and academic perspectives. It would have been useful to have had contributions from people more directly representing views of the private sector.
Bradley J. Olson*
2 See, e.g., the review of Freedman's book infra at 177.
*Mr. Olson received his B.A. (Biology) from Central Connecticut State University and his M.S. from the University of Connecticut. He has several years of laboratory experience in biochemistry and has completed his first year at Franklin Pierce Law Center.
Top of page Index to Risk Book Reviews
Conrad G. Brunk, Lawrence Haworth & Brenda Lee, Value Assumptions in Risk Assessment: A Case Study of the Alachlor Controversy (Wilfred Laurier University Press 1991) [161 pp.] Glossary, notes. C91-094453-9; ISBN: 0-88920-200-1. [Cloth $US 35.00. Waterloo, Ontario, Canada N2L 3C5.]
Readers of RISK should be most interested in the authors' efforts to peel open the classic risk-assessment model and show how such an inquiry lends itself to an infusion of values. "What is at stake in this debate," they say, "are conflicts among fundamental social and moral values, which cannot be resolved by scientific inquiry alone." For example, they cite the Review Board's pervasive allusions to its belief that the manufacturer of the sole competitive product would be left with a monopoly and the effect this belief had on its recommendation.
Further, the authors argue that a vision of technology as the indispensable problem-solver predisposed Board members to select some and ignore other factors for their risk-benefit analysis. Thus, economic gain from improved crop yields was chosen as the benefit to be weighed against farmhand exposure to a potential carcinogen as the risk. Other potential considerations, such as the risk of well-water contamination were not considered, and the authors believe that this could have been expected from an inquiry conducted only by scientists. The selection of factors for analysis and their ultimate weighing necessarily turn on values for which the classic risk-assessment model is ill-suited, warns Value Assumptions.
The book is not without its shortcomings. Perhaps the most serious is its failure to discuss the Pest Control Products Act of 1969 in more than two small paragraphs. The book itself is quite value-laden in its criticism of the Board's process, and lack of attention to the Board's charter leaves one unsure that its process was other than one the legislature dictated. If so, criticism of the Board is misdirected.
"Allow me to frame the issues, and I will control the outcome" is a proposition the truth of which most lawyers quickly realize. In showing how fundamental political, social and moral choices frame and dictate the outcome of a seemingly value neutral process, Brunk, Haworth and Lee make a strong case for wider recognition of the importance of issue framing by scientists -- particularly in situations where scientific knowledge is being used to evaluate and manage risk.
Todd F. Volyn*
*Mr. Volyn received his B.S. (Chemistry) from the University of Louisville (1981) and has experience in the medical diagnostics industry. He is completing his studies at Franklin Pierce Law Center.
Top of page Index to Risk Book Reviews
Warren Freedman, Legal Issues in Biotechnology and Human Reproduction (Quorum Books 1991) [240 pp.] Appendices, bibliography, index to cases, index to subjects. LC: 90-20709; ISBN: 0-89930-635-7. [Cloth $55.00. 88 Post Road West, Westport, CT 06881]
First, he surveys the legal implications of non-traditional approaches to conception and birth. Then, he devotes chapters to topics such as artificial insemination, surrogate motherhood and the role of religion (including Judaism) and morality in the creation of the "new" family. Also, he succinctly discusses such diverse subjects as cryo-preservation of both sperm and egg, in vitro fertilization, sperm and egg banking, patentable new organisms and the ramifications of mapping the human genome. Rights and responsibilities as applied to fetuses, donors and adoptive parents are considered in conjunction with the possible role of physicians, parenthood organizations, researchers, corporations and government. The inadequacies of existing statutes, policies and contracts affecting such issues are reviewed. Mr. Freedman concludes with several proposed solutions.
Insofar as it it is recognized that a child may have as many as five parents -- an egg donor, a sperm donor, a surrogate mother who gestates the fetus and the couple who raise the child -- the permutations are numerous. The author describes myriad legal issues that can arise from these relationships (such as prenatal injuries, wrongful birth and wrongful life actions, abortion issues, paternal versus maternal rights, as well as the tort liability for the sexual transmission of AIDS) and highlights the inadequacy of legal standards to address the rights and responsibilities of all relevant parties.
In a particularly interesting section, at 141-3, Freeman observes:
[T]he eternal conflict between maternal rights and fetal rights has never been resolved, although it appears that courts generally favor fetal over maternal rights. Moral, ethical, and religious opinions intermeddle, and frequently the fetus is protected at the expense of the mother. Undoubtedly a pregnant woman's medical decisions will affect her own life, as well as the life of the fetus she is carrying; her right to her bodily integrity may have to succumb to the state's compelling interest in the child after the first trimester of her pregnancy. On the other hand, the mother does have a legal right to refuse medical treatment pursuant to the informed consent doctrine, provided that the fetus is not viable.In light of the controversy surrounding Roe v. Wade, fetal tissue research, homosexual marriages and AIDS, this book is timely. Anyone interested in the resolution of these and important related issues should find it valuable. Also, because it is well documented, students in any number of disciplines will find it a good place to begin research on a wide variety of human reproductive issues, from both domestic and international perspectives.Here, he also discusses matters such as workplace discrimination and "fetal abuse" that pose many questions as yet answered.
Mark G. Bloom*
*Mr. Bloom received his B.S. (Microbiology) from Ohio State University and his J.D. from Franklin Pierce Law Center. He is associated with Dunn & Simpson, P.C., in Buffalo, NY.
Top of page Index to Risk Book Reviews
Thomas Gilovich, How We Know What Isn't So: The Falliability of Reason in Everyday Life (The Free Press 1991) [216 pp.] Index, notes. CIP: 90-26727; ISBN: 0-02-911705-4. [Cloth $19.95. 866 Third Ave. New York, NY 10022.]
Its author, a professor of psychology at Cornell, explains some of the reasons people often hold beliefs erroneously thought to have empirical support. Readability is enhanced by many examples drawn from common experience; e.g., anyone who follows basketball will enjoy the author's discussion of the "hot hand" myth. Another interesting myth is the "Sports Illustrated jinx."
A major explanation for such beliefs is incomplete or erroneous information. The "hot hand" myth seems to arise in part because people are unaware that, in random series, sequences of identical events are to be expected. The jinx myth arises because many people are unaware of what regression to the mean implies for a sports figure in the midst of an extraordinarily good season. However, other erroneous beliefs are linked to the media's (and other people's) tendency to omit boring details or to slant a report to make an interesting story.1
The second major explanation is lack of adequate rigor; e.g., in assessing the predictive value of dreams, people often ignore non-confirming events. Closely related is failure to control for alternative causes.
Gilovich argues that such errors would be less common if more people were exposed to the scientific method. In the final part (one chapter), he argues specifically for more and better education in statistics and social sciences and offers strong support for this proposition.
For readers of Risk, particularly educators, the following included quotation (and the study from which it was taken) would seem to have important implications:2
It appears that the probabilistic sciences of psychology and medicine teach their students to apply statistical and methodological rules to both scientific and everyday-life problems, whereas the nonprobabilistic science of chemistry and the nonscientific discipline of law do not affect their students in these respects.I will leave it to others to explore possible limitations of the study. Yet, in view of my experience in undergraduate chemistry3 and considerably more extensive experience in legal education, its conclusions do not surprise me. Moreover, given the roles that lawyers and chemists -- not to mention other physical scientists and engineers -- often play in preventing and redressing health and safety risks, those conclusions provide cause for concern.
...
[T]he luxury of not being confronted with messy problems that contain substantial uncertainty and a tangled web of causes means that chemistry does not teach some rules that are relevant to everyday life.
While little of Gilovich's book bears directly on managing health and safety risks, readers interested in how we, both as laypersons and experts, "know what isn't so" will still find much of value.
Thomas G. Field, Jr.
2 Lehman, Lempert & Nisbett, The Effects of Graduate Training on Reasoning: Formal Discipline and Thinking about Everyday-Life Events, 43 Am. Psychologist 431, at 438 and 441 (1988).
3 Chemists' training was found to be even less useful than lawyers' for successful performance of evaluated "reasoning" tasks!
Top of page Index to Risk Book Reviews
Peter W. Huber, Gallileo's Revenge: Junk Science in the Courtroom (Basic Books 1991) [274 pp.] Acknowledgements, index, notes. CIP: 91-70055; ISBN: 0-465-02623-0. [Cloth $23.00 Harper & Row, 10 E. 53d St., New York, NY 10022.]
No reasonable person would expect juries always to be "right." As discussed in an earlier issue of Risk, the jury system is not perfect, but it is the best we have thus far been able to devise.2 Even Huber does not seem to contemplate replacing juries with, e.g., panels of scientists. Rather, based on his carefully selected and anecdotal evidence, he proposes reform of rules designed to keep irrelevant, immaterial and inflammatory evidence away from juries.
Historically, the rule having the most to do with scientific evidence has been the Frye test.3 Basically, it excludes expert testimony based upon a theory or technique, e.g., polygraph evidence, failing to have "general acceptance in the particular field to which it belongs."4 There are many problems with this rule. For example, to what field does polygraph analysis or, for that matter, phrenology belong?5 Thus, in many jurisdictions the Frye rule has been replaced by other rules designed to keep pseudoscience from jury consideration.
Huber apparently believes that these rules are inadequate Yet, he does not advance a specific proposal. Resurrection of the Frye rule, alone, would not seem to accomplish what he seeks. To recover for a compensable injury, a plaintiff need only demonstrate that it is more likely than not that the defendant caused it. This standard of proof is far lower than the 95% confidence level minimally required for most "scientific" purposes and accounts for at least some of the verdicts he criticizes. If such a high standard of proof were adopted, few defendants would be found liable, and the consequences could be worse than those Huber reports.
Indeed, something approximating that level of proof (beyond a reasonable doubt) is required to convict a defendant in a criminal case. Giving that particular jury some benefit of the doubt, the application of such a standard may have resulted in acquittals in the recent police brutality trial in Los Angeles. Fortunately, people rarely react as violently to decisions with which they disagree as people did there. Yet, the situation starkly illustrates the consequences of wide-spread lack of confidence in the legal system.
To avoid a cure potentially worse than the disease, Huber and others who bewail the impact of pseudoscience on jury verdicts should seek more effective strategies for demonstrating that such "science" is the junk they believe it to be. If nothing else, such an approach is far less patronizing of citizens who respond to calls for jury duty.
Thomas G. Field, Jr.
2 Bownes, Should Trial by Jury be Eliminated in Complex Caes? 1 Risk 75 (1990). (Judge Bownes was appointed to the U.S. Court of Appeals in 1977, after eighteen years experience as a lawyer and eleven as a trial judge.)
3 From Frye v. U.S., 293 F.2d 1013 (1923).
4 Id., at 1014.
5 For a recent analysis of Frye and related rules (by a Ph.D. chemist and lawyer), see Grossman, Science and Scientific Evidence, 32 Idea 179 (1992).
Top of page Index to Risk Book Reviews
Stephen Klaidman, Health in the Headlines: The Stories Behind the Stories (Oxford University Press 1991) [249 pp.] Bibliographical references, index. LC: 90-19297; ISBN 0-19-505298-6. [Cloth $24.95. 200 Madison Ave., New York, NY 10016.]
This is an examination, by a well-known journalist and current fellow at the Kennedy Institute of Ethics, of the impact of various media upon public perceptions of the risks posed by EDB, radon, AIDS, cholesterol, the greenhouse effect, nuclear power and cigarettes. With these examples, Klaidman describes and analyzes how media "events" are born, expand into the public's conscience and ultimately affect private and public attempts to cope.
Throughout, Klaidman describes how health risks, of often uncertain severity and frequency, eventually catch the attention of the average citizen. In doing so, he shows how many important technical details are filtered out by the media because journalists find scientific studies difficult to understand -- particularly when they offer conflicting views of the level or existence of risk. He also shows, for example, how distortion can arise from journalists' searches for stories with high drama and human interest.
In some cases, he believes that risk has been overstated. For example, in the chapter on nuclear power, he contrasts the lack of coverage of an allegedly "inherently safe reactor," available since 1967, with the extensive coverage of Three Mile Island. At 99, he speculates that "it could easily be that if the press had given more coverage to the 'inherently safe reactor' twenty years ago, this country's energy grid would look different from the way it looks today."
Yet, he believes that the risk posed by radon has been understated, and his discussion is not without drama. It begins with an account of a Pennsylvania nuclear engineer who inexplicably triggered alarms at the power plant where he worked and later was found to have become "hot" from the radon gas in his home!
In addition to biases introduced by journalists' lack of technical background and their search for drama, Klaidman describes how product manufacturers, political candidates and citizen activists attempt to manipulate the media for their own purposes. The result may be unwise regulation when officials react to snowballing public pressure caused by fears generated by press speculation. In other cases, the responses of individuals may have, in the aggregate, equally or more serious consequences. Consider the following, at 172:
The Associated Press, for example, put a story out that noted: "Adding a bran muffin or a bowl of oat-bran cereal to a daily diet can cost about $250 a year, compared with some cholesterol-lowering drugs, which can cost as much as $1450 a year." What the story didn't say was that the drugs will lower your cholesterol; a muffin or a bowl of cereal will not. [Endnote omitted.]Klaidman's account is anecdotal and aimed at the general public. In some instances, it is apt to spark disagreement. Yet, it is lucid and well documented. Insofar as his book allows readers to consult original sources, many readers of RISK should find it personally interesting and professionally useful.
Eugene Shkurko*
*Mr. Shkurko received his M. Eng. (Electrical) from S.U.N.Y. at Buffalo and his J.D. from Franklin Pierce Law Center. He is a patent attorney with IBM.
Top of page Index to Risk Book Reviews
Sheldon Krimsky, Biotechnics and Society: The Rise of Industrial Genetics (Praeger 1991) [280 pp.] Bibliography, figures, index, list of acronyms used, tables. LC: 90-23214, ISBN: 0-275-93860-3. [Paper $17.95. P.O.B. 5007, Westport CT 06881.]
An analysis of technological change is usually made with the benefit of hindsight. This book provides a unique and interesting view from within an ongoing technological revolution. Drawing on his background, Krimsky documents present concerns with regard to biotechnological research and advancements. His purpose in so doing is to preserve freshness and skepticism with regard to the subject matter and to provide future generations with a record of the questions being asked, the options available and the decisions being made in the field at this time.
The analysis begins with an overview of the industrial context of biotechnology. Krimsky traces developments from the advent of microbiology in the 1700's to the "biotechnics" of today. This part of the book includes a number of informative graphs, charts and tables. An entire chapter is dedicated to the role of patents in biotechnological advancement and makes the important point that a patent does not grant any right to use or sell a patented invention, only to exclude others from making, using, or selling it. Here, the capacity to patent living organisms is treated as raising issues distinct from others concerning genetic engineering, with public perceptions and industrial implications being featured.
Krimsky then gives an in-depth analysis of the relationship between academia, corporations and government with regard to biotechnology before proceeding to a discussion of genetics and ecology. Frost-reducing bacteria are used to exemplify the problems faced when moving genetic research from laboratory to field tests, including public reaction to risk potential.
Chapters in the final section of the book deal with current social controls for regulating new biotechnologies. A discussion of several fragmented programs illustrate what Krimsky believes to be a significant lag in the development of methods to assess advances in biotechnology that underlie new applications. He argues that methods for selecting the most appropriate applications of new developments are critical and that emphasis needs to switch from incentives to develop new applications to incentives encouraging assessment and selection of new basic developments as they are made.
Biotechnics presents a well-written overview of a rapidly developing area of technology and the concerns it has generated. Social, legal, ethical and ecological problems abound, as do governmental programs designed to deal with them. Because such topics are of wide concern and are treated comprehensively, this book should have broad appeal.
Allan M. Wheatcraft*
*Mr. Wheatcraft received his B.S. (Chemical Engineering) from the University of Virginia and has experience in the polymer industry. He has completed his second year at Franklin Pierce Law Center.
Top of page Index to Risk Book Reviews
Elaine Vaughan, Some Factors Influencing the Nonexpert's Perception and Evaluation of Environmental Risks (Garland Pub. Co. 1990) [249 pp.] Abstract, appendices, bibliography, figures, tables. LC: 90-13855, ISBN: 0-8420-0422-1. [Cloth $59.00. 136 Madison Ave., New York, NY 10016.]
Vaughn begins, at 1, by noting that "Most studies [of those factors]... have relied on survey data and correlational analyses... that preclude the determination of interactive effects, effects that could explain apparent `inconsistencies' that have been documented." She therefore set out to empirically test the effects of six variables: "(1) familiarity with the terms used to describe a hazard, (2) environmental persistence of a chemical, (3) personal relevance of data used to evaluate cancer-causing potential, (4) personal relevance of possible adverse consequences, (5) perceived control over exposure and (6) vividness of the exposure pathway."
In the first 50 pages, Vaughn extensively reviews the literature dealing with perception of risk. A large number of variables may have relevance to this topic, and she reviews many of them, including: personal relevance, reversibility of consequences, characteristics of exposure, familiarity, personal control, vividness, decision making for self and others, perception of hazard affecting self and others, and situational influences on perception of control.
Vaughan then describes three studies she designed and executed to determine whether the six experimenter-controlled variables listed above have an effect upon judgments by nonexperts of perceived risk, acceptability of risk, subjective probability of negative outcomes due to exposure to a substance and perceived severity of consequences. Her methodology involved fairly standard procedures: presentation to the subjects of manipulated descriptions of chemical substances (chemical hazard profiles) with questionnaires and rating scales to measure perceptions of risk.
In Chapter 5, Vaughn summarizes her findings, relates them to previous research, and briefly addresses methodological considerations and implications for further research. Those not trained in the social sciences will find this chapter of most interest, but the discussion is likely to be more useful to professionals doing the same kind of research. The interactions among the variables Vaughan studied are complicated -- so complicated, in fact, that she fails consistently to call the variables by the same names. Also, insofar as she deals with information outside her primary area of expertise (psychology), the discussion is sometimes confusing.
How much aid this work will provide -- particularly to those outside of psychology -- is uncertain, but Vaughn is to be encouraged. Overall, this work represents an ambitious effort to serve a laudable goal. For experts to gain acceptance of their risk assessments, they must identify relevant parameters and couch their presentations in effective language and forms that nonexperts can deal with.
Perhaps understanding will improve with further refinement and definition of variables of the type that Vaughn has studied. Yet, given the complexity she demonstrates, examination of variables such as cognitive biases and heuristics may prove more fruitful in the long run.
Juanita V. Field*
*Professor Field teaches Psychology at Plymouth State College.
Top of page Index to Risk Book Reviews
Biomedical Politics (Kathi E. Hanna ed., National Academy Press 1991) [352 pp.] Preface and acknowledgements, Carl W. Gottschalk, Chair, Institute of Medicine Committee to Study Biomedical Decision Making. Appendices, biographical notes on authors and commentators, index, notes, references. LC 91-18394, ISBN 0-309-04486-3. [Cloth $29.95. Export $36.00. 2101 Constitution Ave., NW, Washington DC 20418.]
The committee, itself, is also represented in an introduction, twelve commentaries written by committee members (two follow each case study), and a final chapter summarizing committee conclusions and suggestions for further research. Moreover, committee member Stanley Joel Reiser prepared a paper discussing the historical roles of the public and experts in biomedical policy controversies; this appears as an appendix.
Taken together, the studies in this book provide a good sample of the range of decision-making processes that have been used to resolve complex scientific and medical issues once they have been subjected to public scrutiny. The book is interesting and well written -- in language easily understandable by a wide audience. Also, the commentaries were designed to keep the overall presentation of each topic as objective as possible.
Biomedical Politics demonstrates that important biomedical decisions are at least as much influenced by chance, fate and politics as they are by the merits. This book is highly recommended to anyone interested in how public controversies over highly technical subject matter may be resolved -- whether or not the subject matter be, in any sense, biomedical.
Each topic could easily be the subject of a book as long as this one,* but such a span of pages would defeat its purpose. Nevertheless, it is recommended to anyone seeking an even-handed summary of any of the addressed controversies, and those seeking further information will find the reference lists helpful.
Gianna Julian-Arnold*
* See, e.g., reviews of The Code of Codes..., infra at 261 and Risky Business..., infra at 261 [ed.].
* Ms. Julian-Arnold received a B.S. (Chemistry) from Northern Arizona University and a J.D. from Franklin Pierce Law Center [FPLC]. She is presently an Intellectual Property Research Fellow at FPLC and a candidate for the Masters of Intellectual Property.
Top of page Index to Risk Book Reviews
The Code of Codes: Scientific and Social Issues in the Human Genome Project (Daniel J. Kevles & Leroy Hoodeds., Harvard University Press 1992) [397 pp.] Contributors, figures, index, notes, preface, selected bibliography, tables. LC 91-38477, ISBN 0-674-13645-4. [Cloth $29.95. 79 Garden Street, Cambridge MA 02138.]
The first essay should be of wide interest. It describes how initial attempts to apply genetics to humans led in the U.S. to, e.g., sterilization laws1 and later to the Holocaust. This, in turn, led to increasing scientific hostility and so tarred human genetic research that only slowly have these associations been overcome by work, e.g., on sickle cell anemia. It also recounts how, following identification of the structure of DNA and an invention allowing the isolation and the determination of functions of single human genes, it became possible to sequence the entire human genome. One of the more interesting aspects of this chapter, is its discussion of the mid-1980's origins of the project within the Department of Energy and turf battles with the National Institute of Health (NIH) that resulted, in 1991, with NIH receiving the lion's share of funding and the task of human genetic mapping.
Following a chapter on the technological history of gene mapping and sequencing, the second part the book describes what is necessary to determine our "genetic blueprint" with about three billion DNA base pairs. Three discrete phases are needed to reach that goal: breaking all of the chromosomal DNA into fragments for physical mapping, sequencing of all base pairs of the chromosomes, and, finally, understanding the function of each gene. It is estimated that these phases will require, respectively, one hundred, three to ten thousand, and one million person-years each.
The second part concludes with an essay by the initial project director, James Watson, in which he extols the project's potential for understanding and curing conditions such as Alzheimer's disease, cystic fibrosis and alcoholism -- as well as its cost-effectiveness compared to the cost of the developing palliatives. Watson also expresses his inability to reconcile patenting of DNA sequences with world-wide needs. Ultimately, of course, patenting DNA sequences became the center of NIH controversy and sparked his resignation.2
Watson's essay thus serves as a good lead into a final set of papers that explore potential ethical and social consequences of the genome project. As those papers reveal, the project has potential not only to yield technological solutions for pressing problems but also to outstrip our ability to regulate the technology. For example, it does not take much imagination to envision a genetic underclass created by our capacity to predict but not prevent various conditions, its members being, e.g., shunned by insurers, employers and others.3
Although papers tend to repeat basic information, this well-documented book is highly recommended for its introduction to the technical and social aspects of an important and rapidly-evolving field.
Bradley J. Olson*
2 See, e.g., Leslie Roberts, Why Watson Quit as Project Head, 256 Science 301 (1992).
Ironically, Dr. J. Craig Venter, having a technique for identifying thousands of genes per month, subsequently left NIH to set up his own laboratory. See, e.g., Gina Kolata, Biologist's Speedy Gene Method Scares Peers But Gains Backer, New York Times, July 28, 1992, at C1, col. 2.
3 See, e.g., the review of Risky Business..., infra at 267 (ed.).
* Mr. Olson received his B.A. (Biology) from Central Connecticut State University and his M.S. from the University of Connecticut. He has several years of biochemistry laboratory experience and is a candidate for J.D. and Masters of Intellectual Property degrees at Franklin Pierce Law Center.
Top of page Index to Risk Book Reviews
Roger Cooke, Experts in Uncertainty: Opinion and Subjective Probability in Science (Oxford University Press 1991) [321 pp.] Index of names, index of subjects, tables. LC 90-22493; ISBN 0-19-506465-8 (cloth $65.00). [200 Madison Ave., New York NY 10016.]
Three of the thorniest guardianship questions in probabilistic risk assessment are the focus of Cooke's excellent volume. What is the proper role of scientific experts who advise policymakers regarding existing and proposed public risks? What is the appropriate behavior of experts, decisionmakers and laypersons in societally risky situations characterized by probabilistic or scientific uncertainty? More specifically, what are the ways in which expert subjective probability assessments can either further or thwart rational consensus about public policy?
Ever since the 1979 Lewis Report -- an evaluation of the major probabilistic risk assessment of commercial nuclear fission -- officials have recognized and sanctioned the use of experts' subjective probabilities in risk assessments.1 Yet, some uses of subjective probabilities present obvious and serious threats to the public good and to rational consensus. No one, however, either in the Lewis Report or elsewhere, has systematically and methodically addressed how subjective probabilities ought to be used in estimating and evaluating societal risks. This is the accomplishment of the Cooke volume.
Organized into three parts, the first section of Experts In Uncertainty surveys how risk assessors and policymakers have used expert opinion. The first chapter discusses the Delphi method and scenario analysis, the two main forms in which structured expert opinion was conveyed to decisiomakers during the 1940's, 50's and 60's. The next chapter surveys four important applications of expert opinion: to the aerospace industry, to military intelligence, to the commercial nuclear industry and other objects of probabilistic risk analysis, and to policy analysis. Chapter 3 discusses the emergence, during the 1970's, of expert (artificial intelligence) systems for modeling scientific reasoning under uncertainty.
Chapter 4 summarizes the assets and liabilities of using experts' probabilistic representations of uncertainty. Their chief asset is providing clear criteria for evaluating subjective probability assessments. Their main liability is that, although training in reasoning with uncertainty can be worthwhile, experts typically do not handle subjective probabilities with much skill. Hence, their proffered risk estimates -- of everything from a nuclear core melt to the incidence of pesticide-induced cancer -- are often highly erroneous. To counter the errors exhibited in experts' opinions, in Chapter 5 Cooke provides some guidelines for employing subjective risk estimates. He argues that we need to develop methodological rules for collecting and evaluating subjective probabilities for things such as technological risks.
In Part II, Cooke assembles the mathematical-modeling tools and proofs that will be needed to provide a more suitable method for avoiding errors in the use of expert opinion. He reviews Savage's normative decision theory (Chapter 6), De Finetti's representation theorem (Chapter 7), techniques for eliciting, scoring, evaluating and weighing various experts' probability assessments (Chapters 8 and 9), and calibrating experts on the basis of their past predictive successes (Chapter 10). Cooke argues that, by using these tools, one can improve rational decisionmaking by quantifying experts' uncertainty as subjective probabilities.
Part III of the volume develops three models (classical, Bayesian and psychological scaling) for combining expert opinions into a probability distribution (Chapters 11-14) and evaluating them. It also surveys the experimental results obtained from the author's using these three models in actual technological, industrial and environmental applications. Each of the applications is employed to quantify and evaluate expert opinion about probabilities associated with risks such as space flight, groundwater transport and chemical-plant failures (Chapter 15).
Cooke's groundbreaking work is of great importance to policymakers, risk assessors and moral philosophers because it "downloads" techniques from probability and statistics onto the problem of evaluating the subjective probabilities of risk assessors. The author skillfully uses mathematical tools to help solve a recurrent problem of politics, law and ethics: how to evaluate the decisions of those who, in a democracy, control science, technology and safety.
Another asset of the volume is its clarity of exposition and its quantitative precision. (The author wisely places more mathematical materials in a 25-page appendix.) Cooke's work displays an impressive encyclopedic knowledge: mathematical, scientific and philosophical sophistication combined with first-hand knowledge of how to evaluate the subjective probabilities used to solve real-world problems. His analysis displays his analytic skills as well as his years of experience in applying his methods to the solution of problems in scientific, technological, industrial and environmental risk assessment.
Cooke is especially to be commended for his noting the shortcomings of his proffered models, for example, the strong modeling assumptions underlying the transformations used in the psychological scaling models.2 He also is careful to point out possible future research that would improve his models, for example, the need for parametric techniques to elicit subjective probability distributions.3
The book is organized well, so that those interested in understanding the policy problems associated with the use of subjective probabilities and quantitative risk assessment can read Parts I and III. However, those who wish merely to use Cooke's models in combining and evaluating expert opinions, or in applying them to real-world problems, can read Parts II and III of the volume. In any case, the book is a "must read" for risk assessors, decision theorists, policymakers, mathematicians, attorneys and environmentalists, as well as moral philosophers.
Although there are no major shortcomings in the book, a few readers may wish that Cooke had spent more time tracing the ethical presuppositions and the policy consequences of the three models that he develops for evaluating expert opinion. Because Cooke has come so far in solving a practical problem of politics, law and ethics, however, perhaps it is just as well that he has left most of the evaluation of his solutions to those who follow after him. He has provided a good first step in freeing us from the subjectivity often associated with subjective probabilities in risk assessment.
Kristin S. Shrader-Frechette*
2 At 271.
3 At 270.
* Distinguished Research Professor, Department of Philosophy and Center for Urban Ecology, University of South Florida.
Top of page Index to Risk Book Reviews
Elaine Draper, Risky Business: Genetic Testing and Exclusionary Practices in the Hazardous Work-Place (Cambridge University Press 1991) [315 pp.] Index of names and subjects, glossary, notes, references. LC 90-28112; ISBN 0-521-37027-2 (cloth $49.50); ISBN 0-42248-5 (paper $15.95). [40 W. 20th St., New York NY 10011.]
According to a second view, workplace exposure to certain substances represents a threat only to particular persons. Hence, in order to protect them, susceptible individuals ought to be removed from occupations involving such exposures. On this second view, exposure to chemicals is not generally harmful, since only hypersensitive individuals are affected.
Proponents of the first view tend to blame industry for general chemical hazards to workers and the public. Proponents of the second view, however, tend to blame susceptible workers, particular individuals who are the most obvious victims of chemical exposures. The main purpose of Risky Business is to argue against the second view and therefore against the practice of genetic screening to determine worker susceptibility. Instead, sociologist Elaine Draper argues for stricter health standards for chemical exposures. Like canaries used by coal miners for early detection of methane, Draper claims that genetically susceptible members of high-risk groups also provide warning signals for conditions likely to be dangerous for all workers. If employers react to the warnings by removing the canaries, the high-risk workers, instead of reducing hazardous workplace exposures, Draper argues that deadly consequences will follow for everyone. Less susceptible workers will lose their signals that warned of danger, and exposures and health hazards will increase, ultimately harming even more persons.
Using case studies of two large chemical companies, Dow and DuPont, Draper chronicles how contemporary industries are removing the canaries, the high-risk workers. They also deliberately fail to inform more susceptible employees of possible hazards, deny the risks involved and therefore accelerate the threat to other workers. More importantly, Draper reveals how industry paints a deadly, occupationally induced hazard as a problem of unique, individual susceptibility. Using 120 interviews, as well as documentary materials and workplace observations, Draper provides an account that is sure to become a classic in the risk literature.
Risky Business includes seven chapters as well as 100 pages of references and notes. The first chapter outlines the emergence of genetic testing and susceptibility explanations, surveying the conflict between labor and management over how to define workers at risk from hazardous chemicals.1 Chapter 2 explains the way that industry often defines and debates workplace risk, in terms of specific, genetic propensities of individuals, rather than as a societal threat to virtually everyone. In other words, industry typically deals with workplace risk by "blaming the victim"2 and by projecting social ills onto individuals3 whom industry then defines as unusual or atypical.4 Chapter 3 shows how managers and scientists who accept the individual-susceptibility approach make it appear that all workers and exposure conditions not considered "high risk" are safe.5 In Chapter 4, Draper analyzes the stratification of the work force. She argues that, contrary to its proponents, the new genetic screening technology is neither ethically nor politically neutral. Instead the process often singles out women and minorities.6 Hence, she argues, the search for workers most at risk often results in the identification of inappropriate biological categories.
Chapter 5 surveys the use of genetic susceptibility tests, showing that the results are typically denied to workers, even after they ask,7 and that the tests are rarely used to monitor and improve worker health. Instead, the chapter argues that the results are used largely to "remove the canary" and to deny that any health threat exists. Such denial is possible, Draper argues, because workers in dangerous industries typically are not unionized.8 Chapter 6 uncovers the legal and regulatory environment that is conducive to removing particular employees, rather than to reducing workplace exposures to hazardous substances. Such an environment encourages initial screening, not continual surveillance -- and secrecy, not honesty, about hazards.9 Hence, Draper argues that there is virtually no sense in which genetic screening has helped workers in any way. In Chapter 7, Draper summarizes the arguments and suggests the policy implications of her findings. With the rapid growth of genetic knowledge and the increasing influence of economic, legal and regulatory factors, Draper predicts that exclusionary testing will become even more prevalent in the future.
Draper's book deserves high praise, not only because of the wealth of information about chemical risks, genetics, occupational medicine and corporate policies and politics that it includes, but also because of her compelling sociological thesis. This thesis is that industry assimilates scientific and technological developments -- like genetic testing -- in ways that both reflect and serve its social biases and power relationships. Because her volume attempts to tell a story of causes and motives, and not merely actions and consequences, Draper is likely to be both praised for her insights but attacked by those who are uncomfortable with her case-study methodology. The complex inferences associated with the sociology of risk are typically not amenable to the precise and quantitative techniques of hypothesis deduction often used in other sciences. Because they are not, some researchers may argue that Draper should have done more to build her case with a variety of statistical methods and tests.
In response, however, Draper's defenders should point out that, in order to deal with some of issues she confronts, Draper could have used only informal methods such as interviews, observations and analysis of documents. Also, one book cannot do everything, and Draper's volume is meant to achieve sociological analysis, not statistical proof. In laying a plausible sociological foundation for a compellingly correct thesis, Draper accomplishes more than most authors. She shows us how industry apparently uses advances in science and technology to serve its own economic ends and social biases.
In laying a sociological foundation for future work on occupational risks and genetic screening, Draper has achieved a number of goals that make her work unique and particularly noteworthy. For one thing, the type of sociological analysis that she uses has never before been brought to bear on the question of exclusionary testing of workers in the chemical industry. Second, Draper consistently employs a detailed, analytic treatment of alternative positions on particular issues related to genetic testing in the workplace. She carefully exposes the assumptions and traces the consequences of opposing arguments, revealing logical and social analysis at its best. Third, Draper does a superb job of showing how the shift from emphasis on chemical hazards to emphasis on individual susceptibility is neither a scientific breakthrough nor an objective analysis of recent consequences of genetic developments. Rather, she reveals that the shift is a socially constructed account designed to serve special interests. Fourth, Draper serves policymakers well by outlining specific alternatives to exclusionary genetic testing in the workplace. For all these reasons, her volume is a landmark discussion of the sociological study of workplace risk.
Kristin S. Shrader-Frechette*
2 At 37.
3 At 39.
4 At 44, 49.
5 At 59.
6 At 75-83.
7 At 100.
8 At 114.
9 At 157.
* See previous review.
Top of page Index to Risk Book Reviews
Leonard Evans, Traffic Safety and the Driver (Van Nostrand Reinhold 1991) [404 pp.] About the author, author index, figures, notes, preface, subject index, tables. LC 90-44233, ISBN 0-4421-00163-0. [Cloth $51.95. P.O.B. 668, Florence KY 41022-0668.]
Deaths, injuries and property damages from car crashes -- their origin and nature, and ways to prevent their occurrence and reduce their severity -- form the subject of this book. Traffic crashes are perhaps the number one public health problem...; almost half of the 19-year-olds who die do so as the result of traffic crashes. ...From there, Evans goes on to analyze and synthesize a vast body of research. Some of it is widely known, e.g.: motor cycle helmets and seat belts, when used, reduce the severity of injuries in crashes, age and sex affect the likelihood of a crash; drivers and passengers tend to be safer in larger cars than in smaller ones; and alcohol consumption is the most significant cause of traffic collisions. Other data is less well known, e.g., that risk of injury to people in cars struck by another vehicle is inversely proportional to the weight of their car, the effects of driver personality on crash frequency, the makes of cars involved in the most crashes, and the effectiveness of safety devices and drunk driving countermeasures.
Because at least one driver is involved in every traffic crash, there is considerable focus on the driver.
Most early chapters discuss factors that affect the likelihood and consequences of vehicle crashes, but Chapter 3 presents a considerable amount of basic data on the frequency and severity of traffic fatalities in the U.S. Later chapters attempt to assess the relative importance of factors and examine traffic safety in broader contexts. Chapter 12, entitled, How You Can Reduce Your Risk, is likely to be of wide interest.
Chapters that present research results generally omit details concerning the methods used to obtain them but present conclusions drawn by the author. For example, in Chapter 9, entitled, Effectiveness of Occupant Protection Devices When They Are Used, Evans concludes in part:2
Nearly all of the lap belt effectiveness, and almost half of the lap/shoulder belt effectiveness, is due to the prevention of ejection.While the book's strength is its coherent presentation of hard data, the solutions suggested by the author are mostly soft and non-technology oriented. Evans suggests that changes in attitudes, particularly in regard to alcohol consumption by drivers, are most likely to improve traffic safety.
By assuming that airbags do not influence ejection risk, and that they provide interior impact reduction effectiveness equal to that of lap/shoulder belts, effectiveness is estimated at 17%.... All of these results are for fatalities only, and cannot be extrapolated to lower levels of injury.
Given Evans' position, the certainty that anything he might say of possible benefit to plaintiffs' attorneys would be used in crash-worthiness3 suits against GM and other auto manufacturers, and the obvious impact of the research he presents on, e.g., regulatory requirements for additional safety-related hardware, this is not surprising. Nevertheless, its comprehensiveness and readability, coupled with its extensive references to the primary literature, make this book an excellent starting place for anyone interested in traffic safety research.4
Allan M. Wheatcraft*
1 Preface, at xiii.
2 At 247-8.
3 See, e.g., 15 U.S.C. § 1901(14) (1982); see also, e.g., Evans v. General Motors Corp., 359 F.2d 822 (7th Cir. 1966), cert. denied 385 U.S. 836 (1966).
4 See also, e.g., John D. Graham, The Safety Risks of Proposed Fuel Economy Legislation, 3 Risk 95 (1992).
* Mr. Wheatcraft received his B.S. (Chemical Engineering) from the University of Virginia and has experience in the polymer industry. He is completing his J.D. program at Franklin Pierce Law Center.
Top of page Index to Risk Book Reviews
Iain Hay, Money, Medicine and Malpractice in American Society (Praeger 1992) [244 pp.] Abbreviations, annotated list of personal communications, figures, glossary, index, full legal citations, notes, preface, references, tables. LC 91-38477, ISBN 0-674-13645-4. [Cloth $49.95. P.O. Box 5007, Westwood CT 06881-9990.]
Those interested in the topic will find Money, Medicine and Malpractice to offer an interesting and relatively broad perspective. Hay, a Lecturer in Geography at the Flinders University of South Australia, appears to be neither, e.g., doctor, lawyer nor "American" (a term offensive to those living south of the Rio Grande). Yet, the book gains depth from his earlier work in writing The Caring Commodity: The Provision of Health Care in New Zealand (1989).
Hay begins with the observation that:2
The "American Way" is incompatible with the U.S. experience of post-World War II capitalism. National and individual self-determination are collapsing in the face of profit-seeking, social compulsions, and the imperative of global economic competition. Only the illusion of free choice and the misguided rhetoric of individualism remain: they mask new realities of compulsion and collectivism. This cultural contradiction is explored through an investigation of the development of medical liability insurance and its implications for tort reform and health care provision in the United States.Only his treatment of reinsurance -- particularly as influenced by Lloyd's of London and other foreign insurance organizations -- is missing from that summary.3 Hay says that his book can be viewed as examining: (1) "trans-formations in medicine, law, and society that brought about and sustained the growing need for, and importance of, medical liability insurance...," (2) "some of the consequences for the U.S. malpractice market of physicians' dependence on insurance in the new global context of communications, proximity, capital mobility, and economy," and (3) consequences, namely physicians' loss of professional control and patients' potential loss of legal recourse.4 Also, it follows a two part chronological division between the evolution of malpractice during a period of wealth and changes induced by "conditions of growing financial stringency."5
In the first chapter, Hay reports that, unlike the situation in the U.S. or the one emerging in the U.K., a no-fault system in New Zealand and a similar scheme in Sweden are fairly popular.6 Yet, perhaps because he is an outsider, he fails to weave this into his conclusions. Thus, we are left to wonder whether the U.S. might be better off with such a system -- and, if so, why.7
Nevertheless, physicians, lawyers and others interested in malpractice should find this well-produced and extensively-documented account of the interactions of medicine, law and insurance to be of considerable interest.
Thomas G. Field, Jr. 1 Richard Stradling, Malpractice Suit Award Worries Doctors: Man Who Lost Penis Given $3 Million, Concord Monitor, Sept. 5, 1992, at 1.
2 Preface, at xxi.
3 See, e.g., Reinsurance and Shifts in Influence, Chapter 7.
4 Preface at xxiii-xxxiv.
5 Id. at xxxiv.
6 At 13.
7 Along this line, see, e.g., Stephen D. Sugarman, Doing away with Personal Injury Law (1989) -- reviewed in 1 Risk 277 (1990).
Top of page Index to Risk Book Reviews
Helena Chmura Kraemer, Evaluating Medical Tests: Objective and Quantitative Guidelines (Sage Publications 1992). [295 pp.] Foreword A. John Rush, M.D. Acknowledgments, index, list of statistical notations, preface, references, and about the author. ISBN 0-8039-4611-2 (cloth); 0-8039-4612-0 (paper). [Cloth $45.00, paper $22.50. 2455 Teller Road, Newbury Park CA 91320.]
Kraemer has written an excellent book that develops a procedure for evaluating the performance of medical tests. The methodology considers the probabilities of incorrect results, the consequences of such errors, and costs. Kraemer argues correctly that future evaluation of medical tests should be based on systematic, completely objective and empirical methodologies. This book takes a significant step in that direction.
The book would be beneficial for anyone concerned with evaluating medical test performance. However, to profit from it, readers will need a moderately good grasp of basic statistics and of current approaches to assessing medical tests. Thus, this work is essential reading for those with a serious interest in methodologies for assessing medical test quality, and who have or are willing to develop a sound understanding of the field.
In the first five chapters, Kraemer thoughtfully reviews the basics of medical testing procedures. She discusses such topics as disorder vs. diagnosis; testing protocols, responses and referents; and population and sampling concepts. She also considers sensitivity, specificity, predictive values, efficiency, and statistical tests for significance. She devotes substantial discussion to the variability of diagnoses, and constructs an index measuring diagnosis reproducibility. She thereby demonstrates that the outcomes of diagnoses and gold standards do not perfectly identify the presence or absence of a disorder.
In Chapter 6, Kraemer critically analyzes the signal detection approach to test evaluation. She identifies problems with sensitivity and specificity, develops quality coefficients for each to correct the problems, and presents test Receive Operating Characteristic (ROC) curves for sensitivity/specificity and test Quality Receiver Operating Characteristic (QROC) curves for her quality coefficients. She clearly explains ROC and QROC curves, and why tests on them are preferable to those below the curves. She then takes the important step of developing ROC and QROC curves for the diagnosis, that is diagnosis ROC and QROC curves, and demonstrates how the best tests will be those lying closest to or above the diagnosis ROC or QROC curves. Her procedure provides an objective method, based on empirical evidence, for evaluating medical tests, and is a significant contribution.
Also in this chapter, Kraemer shows that a random test will have a sensitivity equal to the probability of a positive test result. (Test sensitivity measures the percentage of those with a disorder who test positive for it.) For example, assume a patient is tested by flipping a fair coin, and classified as positive if a head appears. Then, 50% of those with the disorder will test positive, resulting in an expected sensitivity of 50% by random chance. To correct for this problem, Kraemer develops a sensitivity quality coefficient, which measures the percentage of those with the disorder who test positive after deleting the random element. It equals (Sensitivity - Random Percentage)/(1 - Random Percentage). In the preceding example, where sensitivity was 50% by random chance, the sensitivity quality coefficient would be zero, since (0.50 -0.50)/(1 - 0.50) = 0. Now consider another test, which also has a 50% random percentage, but in which 75% of those diagnosed with the disorder test positive. For this test, the quality coefficient is 50%, since (0.75 - 0.50)/(1 - 0.50) = 0.50. In a third test with a random percentage of 50%, in which all of those diagnosed with the disorder test positive, sensitivity is 100%. And, (1 - 0.5)/(1 - 0.5) = 1, meaning that the quality coefficient is also 100%. Kraemer develops a similar quality index for specificity.
However, the book seems to suggest that the two quality coefficients always are preferable to sensitivity and specificity. This is not so. Assume for example two screening tests for a serious disease that is curable if detected. One test has a random level of 0.20 and a sensitivity of 0.40, while the other has a random level of 0.90 and a sensitivity of 0.91. The sensitivity quality index of the first test is 0.25, and of the second 0.10. For the first test, 40% of those with the disease will be detected, while with the second test, 91% of those with the disease will be detected. Yet, using only the quality indices, the first test incorrectly would be considered the best. My example is extreme and ignores specificity. But, it suggests that there may be cases where sensitivity is a better measure than its quality index, or at least that both should be considered. I don't think that Kraemer would disagree with my comments, and her final model avoids the problem. But, readers might draw the wrong conclusions about the universal superiority of the quality indices from the Chapter 6 presentation.
In Chapter 7, Kraemer covers the second major approach to test evaluation the Predictive Value, or Bayesian, approach. She argues that there is no fundamental difference between this approach and the signal detection method when sensitivity and specificity are replaced by her quality coefficients for these two terms. And, in Chapter 12, she discusses the third major technique, the Multivariate Discrimination approach. In both cases she provides excellent brief descriptions and critiques of the procedures.
Another contribution to test evaluation is covered thoroughly in Chapter 9. Here, Kraemer incorporates into her model the utilities of correct and incorrect test results, and the cost of the test itself. By the end of this chapter, she provides a performance index for medical test evaluation that includes test accuracy, relative clinical benefit, fixed test cost, and potential clinical benefit. Because this final model is logically satisfying and requires data that normally would be available, it is of practical value.
A further significant contribution is Kraemer's development of a procedure for evaluating batteries of medical tests by using optimal sequences, covered primarily in Chapter 11. She also discusses prognostic tests, and concludes with a summary of the past, present and future of medical testing. In her final chapter, she argues convincingly for a standardized procedure for test evaluation, with mandatory requirements along the lines of the standardization provided by the Food and Drug Administration for the evaluation of new drugs.
In summary, this book provides a unique and thought-provoking examination of medical test evaluation, illustrating what is wrong with many current methods. By offering a consistent and logical alternative, it makes an important contribution toward improved evaluations of medical test quality.
Darold T. Barnum*
* Dr. Barnum is a Professor of Management at the University of Illinois at Chicago, and Director of the University of Illinois Center for Human Resource Management.
Top of page Index to Risk Book Reviews
Risk-Taking Behavior (J. Frank Yates ed, Wiley 1992) [244 pp.] Acknowledgements, author index, figures, preface, references, series preface, subject index, tables. LC 91-21229, ISBN 0-471-92250-1. [Cloth $64.95. 1 Wiley Drive, Somerset NJ 08875.]
In the first chapter, the editor attempts to identify consensus within the far-flung, multi-disciplinary research on risk and concludes:1
At its core, risk is the possibility of loss. There are three essential risk elements: (a) losses, (b) the significance of those losses, and (c) uncertainty associated with those losses. ... However, it is not entirely clear that the various acknowledged elements of risk can be consistently aggregated to form overall risk.This summary seems to verify immaturity of knowledge of the topic; consensus occurs only at the most superficial level, and there is little agreement about how to proceed from that point. As a result, the several papers depart in widely varied directions, each with its own extensive discussion of theory and thorough summary of research.
The primary emphasis is on theory and, secondarily, use of theory to explain research, but where theories are not available, some authors have settled for the precursor of raising issues. Unfortunately, while theory is expected to simplify, the issue-raising approach often causes further complication.
Yet, this attempt to present many facets of the study of risk in one location makes neglected areas more noticeable. For example, the chapter on developmental aspects of risk-taking behavior assumes a decision-making perspective and relates this to age, beginning with adolescents. Especially within the context of some particular examples, such as drug use, it would be helpful to begin earlier and consider the risk-taking behavior of younger children.
As an example of the coverage of the topic, consider the chapter on individual differences in risk-taking. Bromley and Curley begin by identifying three questions to be answered: (1) Is risk-taking a personality trait? (2) What theories may allow description, explanation and prediction of individual differences? (3) What applications flow from the research results? This is followed with a description of six approaches to the study of risk-taking as a "trait," leading to a discussion of the importance of a situational/trait interaction perspective and the identification of four categories of situations (physical sensation, games and lotteries, everyday life, and business). Finally, cross-situational studies are discussed, and the authors conclude that: "Greater attention to and analysis of situational factors would provide a valuable addition to our understanding of the role of individual differences in risk taking behavior."2
The papers in this book are well-written and readable. Together, they are comprehensive. Touching what seems to be all of the bases, they present an enormous amount of information. The book will probably be of most use to social scientists, especially psychologists, who are already somewhat familiar with the terms, issues and research, but the chapters on optimality and risk appraisal may be more familiar to readers in disciplines other than psychology.
Yet, anyone interested in risk-taking behavior will benefit from the opportunity that this book affords to become familiar with the broad span of literature addressing the topic.
Juanita V. Field*
2 At 126.
* Dr. Field is Assistant Professor of Psychology at Plymouth State College.
Top of page Index to Risk Book Reviews
Thomas Szasz, Our Right to Drugs: The Case for a Free Market (Praeger 1992) [164 pp.] Notes, bibliography, name index, notes, preface, subject index. LC: 91-30378; ISBN: 0-275-94216-3. [Cloth $19.95. P.O. Box 5007, Westwood CT 06881-9990.]
From the founding of the American Colonies until the Civil War, marijuana was an important cash crop.... The colonists, including George Washington grew... "hemp".... [H]ow many people know that hemp, coca, and opium poppy are ordinary plants, understand how they became transformed into dreaded "dangerous drugs," and realize that in losing our rights to them we have surrendered some of our most basic property rights?
Thus Szasz begins1 by asking questions that ought be asked when we face crises with regard to both health care and a "War on Drugs." He finishes by warning that our rights may further diminish if we allow paternalism to expand. For example, Szasz observes that, rather than customers always being right, with regard to prescription drugs, it is the doctor who is always right. More particularly, with regard to psychiatric drugs, "`The Patient is always wrong': The psychiatrist decides what drug the mental patient `needs' and compels him to consume it, by force, if necessary."2 Yet, as Szasz later notes, even doctors have given up important rights. "Physicians cannot prescribe for pain as they used to.... But they have become so accustomed to state control of drugs that they never lay the blame... where it belongs...."3 Thus, on the one hand, physicians gain because limits on over-the-counter medicines helps to assure them a never-ending stream of patients, many capable of self diagnosis and treatment; on the other hand, they lose because their ability to prescribe is compromised by "fear of being apprehended by agents of the American drug-police state...."4
In a democratic society, we recognize, indeed value, individual responsibility. Inherent in freedom of choice is the ability to make the wrong choice and to suffer the consequences. Thus, Szasz clearly favors the right of citizens, e.g., to grow and consume marijuana on their privately owned property. But what of those who supply other consumers? He suggests that we do not, e.g., blame the obesity of fat persons on people who sell them food. Why then do we blame the habits of adult drug users/addicts on those who sell drugs?
Szasz suggests that paternalistic-prohibitory drug laws are accepted by citizens as a result of the distribution of disinformation about drugs and human responsibility. He has attempted to tell the other side. If nothing else, it will help us, as citizens, to evaluate what we are told and to consider carefully the rights we surrender.
Jennifer L. Frizzell*
2 At 18.
3 At 127.
4 At 128.
* Ms. Frizzell received her B.A. (Political Science) from the University of New Hampshire and has federal administrative and regulatory experience in Washington, DC. She is now completing her legal studies at Franklin Pierce Law Center.
Top of page Index to Risk Book Reviews
Mary R. English, Siting Low-Level Radioactive Waste Disposal Facilities: The Public Policy Dilemma (Quorum Books 1992) [278 pp.], Appendices, index, introduction, references. LC-91-42774; ISBN 0-89930-560-1. [$49.95 cloth. One Madison Avenue, New York NY 10010.]
As its subtitle suggests, this work focuses on the conflict among many competing public policy interests and, as the author puts it, the "web of issues" that surround the low level waste (LLW) disposal dilemma. The book begins with a well-written background, discussing the legislative history of the Low Level Radioactive Waste Policy Act of 1980 (LLWPA) and subsequent state actions required by the act. The author keenly notes that a key change in the LLW disposal problem was a shift from the NIMBY syndrome (not in my backyard) to a NIABY philosophy (not in anyone's backyard). The LLWPA proposed a deceptively simple solution to this problem; each state will be responsible for its own waste. This first chapter so succinctly covers the legal, technical and political factors involved in this rule, one wonders what the remainder of the book has left to discuss.
However, the author then plunges into a series of deeper social factors that underlie everyday problems. These are: (1) the authority required to implement a waste handling program, (2) the trust that such authority requires, (3) risk analysis, (4) the justice requirement, and (5) the need for legitimacy in any final solution. Each chapter includes an extremely detailed case study to illustrate that chapter's point and to bring it somewhat down to earth. FInally, a significant portion of the book is devoted to appendices providing the full text of the LLWPA as originally passed, its 1985 amendments and a synopsis of a number of state's facility siting programs.
The overall thesis appears to be that more than the superficial legal, technical and political factors must be involved when siting LLW facilities. The author makes a compelling argument that certain social factors must be considered by siting committees and legislatures along with the technical issues before a site selection plan can be successfully implemented. However, to the policy maker confronted with an actual siting dilemma, the book may appear to be too deep to be practical. Yet, if the discussion is intended for social scientists or theoreticians, its scope may be too narrow. Notwithstanding this dilemma, the author presents something for nearly everyone in this exhaustive discussion of an important, contemporary environmental problem. Most importantly, the issues are handled in such a way that no biases were evident. The author is to be commended for the depth and perceptivity of her work.
Peter C. Christensen*
* Mr. Christensen received a B.S. (Electrical Engineering) from the University of Minnesota, Minneapolis and is completing work for his J.D. at Franklin Pierce Law Center. He has several years of experience in the electric utility industry.
Top of page Index to Risk Book Reviews
Robert H. Blank, Mother and Fetus: Changing Notions of Maternal Responsibility (Greenwood Press 1992) [207 pp.], Appendix, bibliography, figures, index, preface, tables. LC-91-38029; ISBN 0-313-27639-0. [$45.00 cloth. 88 Post Road West, Box 5007, Westport CT 06881.]
Dr. Blank identifies and examines three general factors which have contributed to the perceived societal unrest. First, he looks to the emergence of advanced technologies used in diagnosing and treating fetuses. One of the effects of these advances has been to make the fetus a "person," one who can be seen through ultrasound and whose sex can be determined through amniocentesis. No longer is the fetus a mysterious being primarily perceived as an appendage of the mother. This new perception of the fetus has led to concern that this "person" be assured a chance to start life free of handicaps or disadvantages caused by third parties or its mother.
The second factor identified as a cause of changing societal views is the studies and findings that have enhanced the understanding of the relationship between a mother's behavior and effects on a developing fetus. For example, he presents data on the deleterious effects a mother's alcohol consumption and street drug use has on fetal development. Recent findings have resulted in the imposition of moral, and sometimes legal, maternal obligations not to act in ways known to be detrimental to the fetus.
In addressing the third factor, Dr. Blank reviews a plethora of cases demonstrating a variety of legal intrusions into the historical relationship between the mother and child. It has long been held that children may bring actions against third parties for prenatal injuries, but they have become more prevalent. Still, as Dr. Blank reports, jurisdictions vary widely concerning, e.g., whether a live birth is required before recovery on behalf of a child can be had for prenatal injuries and whether injuries sustained prior to viability are compensable. One of the latest and most controversial developments in prenatal injury is abolition of intrafamily immunity in most jurisdictions, thereby opening the door for a child to sue its parents.
In considering the increased instance of cases against third parties and the abolition of the family immunity doctrine, Dr. Blank concludes, "[t]he manifestation of these trends is likely to alter substantially notions of responsible maternal behavior and in some cases pit the rights of privacy and autonomy of the mother against the rights of a child to be born with a sound mind and body."2 Throughout, we are reminded of the lack of easy ways to choose between the rights of mothers and the well-being of fetuses.
As Dr. Blank states in the preface, "[a]lthough the challenges of maximizing fetal health without unduly constraining the rights of the pregnant woman remain daunting, hopefully this book will contribute to a rational dialogue that leads to enlightened policy." Although his analysis is quite limited, the book directs readers to a large body of information addressing many interesting issues and constitutes an important resource for anyone wanting to study the "changing notions of maternal responsibility."
Ann Marie Wheatcraft*
2 At 89.
* Ms. Wheatcraft received her B.S. (Psychology) from the Pennsylvania State University. She has experience in the mental health field and is completing work for the J.D. at Franklin Pierce Law Center.