Prometheus Reimagined

Front matter

Technology, Environment, and Law in the Twenty-first Century

Albert C. Lin peat A pee

Albert C. Lin examines technologies that will be among the most life-changing and controversial in coming decades: synthetic biology, nanotechnology, human enhancement, and geoengineering. While such technologies promise to address many of humanity’s most serious problems, they also bring environmental and healthrelated risks and uncertainties. Moreover, such technologies can come to dominate global production systems and markets with very little public input or awareness. Existing governance institutions and processes do not adequately address the risks of new technologies, nor do they give much consideration to the concerns of persons affected by those technologies.

Contents

LIST OF ABBREVIATIONS iX

PREFACE AND ACKNOWLEDGMENTS Xi Introduction 1 1. Existing Law and New Technologies 15

2. Biotechnology Emerging Technology Past and Present 48

Chapter I

Our society is divided, if not schizophrenic, in its attitudes toward regulation. The regulation of emerging technologies is no exception. While some Americans support stronger regulatory protection, many others expound the virtues of limited government and decry excessive state control. This latter view stresses the costs of regulation and attributes our nation’s success in large part to the freedom enjoyed by entrepreneurs, innovators, and inventors.’ Indeed, Americans generally have great faith in technology and see it as a powerful solution to many of the challenges we face. This faith is reflected both in the popular embrace of new gadgets and in government policies that promote the development of new technologies through grants, tax breaks, and intellectual property protections.

Notwithstanding these attitudes, many Americans also assume that government is protecting us from the excesses of the market and the risks that new technologies may pose. We expect emerging technologies and the products that incorporate those technologies to be thoroughly tested and reasonably safe. We also expect clean air, clean water, and a healthy environment, and we expect that emerging technologies will not injure these resources. Such expectations, however, often are not met. ‘Toys contaminated with lead, baby bottles tainted by endocrine-disrupting chemicals, and the Deepwater Horizon oil spill are just a few recent instances in which human health and environmental regulatory efforts have fallen short.

‘Technological change does not occur on its own, of course. Technologies are a product of human discovery, choice, and policy.’ Accordingly, managing technology not only is possible but also is a critical social endeavor. An essential step in sound technology management is technology assessment, which attempts to predict the course and consequences of technological development. Using the information generated through

I5 such assessments, society can regulate or reshape a technology or seek to direct its development. This is not to suggest, however, that society can exert complete control over technologies and their effects. Even technologies subject to close scrutiny during development and implementation may give rise to unanticipated consequences. The spread of a technology, moreover, may make its control costly and difficult. Ultimately, technologies may transform societal values and society itself.’

The predominant approach to managing technology has been reactive. In the United States, the production and use of new chemicals, for example, is generally allowed until there is hard evidence indicating significant and adverse health or environmental effects.* Similarly, government oversight of nuclear reactor safety was minimal until the Three Mile Island and Chernobyl accidents,’ and regulators did not seriously address the risks associated with deep-ocean oil drilling until the recent Deepwater Horizon disaster. The development of safety requirements in response to the risks laid bare by such incidents is of course rational and appropriate. Addressing harms after the fact, however, can be problematic. In many instances, the harms that materialize are difficult, if not impossible, to remedy. Responses crafted after the fact often are limited by existing infrastructure, prior commitments, and the policy preferences of powerful vested interests. In addition, such responses tend to target the risks associated with the specific events contributing to a harm, rather than provide for a broad and comprehensive evaluation of the underlying technology.

The widespread and potentially irreversible effects of emerging technologies warrant prompt attention and proactive management. As these technologies generally promise great benefit, knee-jerk moratoriums should be avoided. But these technologies also have the potential to adversely affect critical aspects of people’s lives, and the general public should be kept informed of technological developments and have meaningful opportunities to be involved in their management.

Proactive technology management is not a new concept. Policymakers developed the tools of technology assessment and environmental impact assessment decades ago. However, the implementation of these tools has been generally ineffective and has failed to involve the public in meaningful ways. This chapter explores past technology assessment practices and thus sets the stage for the case studies in subsequent chapters and for considering how society might better manage technology. One option would be to more effectively implement existing tools of technology assessment. ‘Though efforts along these lines have been initiated, their impact is likely to be limited. More fundamental changes are needed to bring about participatory and effective management of emerging technologies.

The discussion in this and subsequent chapters focuses on technology assessment and regulation in the United States. The United States is a leader in technological innovation, and it effectively invented the concepts of technology assessment and environmental assessment. Scientific and technological advances often begin here and then spread abroad. Because much of the current research and development activity in nanotechnology, synthetic biology, and other emerging technologies is occurring in the United States, this country offers a suitable context for considering the challenges of emerging technologies. Moreover, governance standards and techniques adopted by the United States frequently serve as models for other nations. The United States does not have a monopoly on technological innovation or technology management, of course. Efforts by other nations to assess and manage technologies also can be instructive. Furthermore, international dimensions of technology management are increasingly important thanks to globalization and global problems such as climate change. International governance thus merits serious attention as well.

Goals of Technology Assessment: Assessment and Public Engagement

Assessment

The practice of technology assessment arose in the 1970s with the ambition of predicting and analyzing the full range of consequences—social, environmental, and otherwise—of a given technology.’ As then conceived, technology assessment was to involve objective analysis that drew on the natural and social sciences; subjective value judgments were to be left to democratically elected officials.* A basic premise behind technology assessment was the belief that society could use and manage technology effectively through the application of comprehensive rationality—that is, that such assessments could be “precise, value-neutral, and exhaustive of relevant concerns” and could thereby enable fully informed policy decisions.’

Conventional technology assessment presumed, moreover, that technology develops in a linear fashion: Basic research would be followed by applied research and then production. As such, technology assessment focused on the latter part of that process and sought to generate an objective analysis of virtually finished technologies. At this stage, the ability to refashion a technology or to shape it to account for concerns identified in an assessment was often limited. Additionally, in providing technical analysis, technology assessors did little to involve the public. Societal values could presumably be incorporated later, when politically accountable actors made decisions on whether and how to regulate a technology.

‘Technology assessment’s faith in information gathering, rational analysis, and human control was overly optimistic, as was the assumption that societal values could adequately be folded in at the end stages of technology production. ‘Technology development is not linear, and there are limits to our capacity to predict and direct technological futures.'’ But the underlying motivation of developing and managing technologies in a more deliberate and publicly accountable manner remains critical. Recent refinements to technology assessment, discussed later in this chapter, have sought to incorporate assessment earlier in technology development, encourage greater reflectiveness among scientists about the ramifications of their work, and integrate public input into the assessment process.

In addition to these shortcomings of conventional technology assessment, there are further concerns regarding risk assessment, which is a central component of technology assessment.'” Risk assessment is a type of technical analysis that seeks to produce quantitative estimates of the probability and magnitude of potential harms from an occurrence.” That occurrence might involve a commonplace activity, such as driving a car, or a complex policy decision, such as the adoption of a new technology. Risk assessment for a new chemical substance, for example, consists of a process to identify potential hazards, estimate the probability and magnitude of injury at different exposure levels, and analyze the likelihood of exposure. The risk assessment process ultimately generates a risk characterization, which ideally includes a range of estimates to quantify identified hazards. A risk characterization may also include discussion of uncertainties, underlying assumptions, and the degree of confidence with which estimates are made.

Risk assessment and other technical analyses cannot provide all the data necessary to make perfectly informed decisions, however. Risk assessment can identify some hazards, but the quantification of risk often involves rough probability estimates whose confidence intervals may be so wide as to render such risk quantification useless.'* Certain hazards, moreover, simply cannot be quantified because insufficient data exist or the ability to perform useful experiments is limited." Beyond that, other hazards—“unknown unknowns”—cannot be identified because of limitations in modern scientific understanding, random processes, and the inherent unpredictability of interactions among technology, society, and the environment.'® For emerging technologies, unknown unknowns loom large because we are often dealing with novel mechanisms beyond our expertise. Decisions regarding technology nonetheless must take uncertainties and unknowns into account, even if these indefinites cannot be quantified or readily described.

These points lead to a more fundamental criticism of risk assessment and by extension technology assessment. The quantitative analyses typically generated by risk assessments tend to hinder the consideration of qualitative factors and other pertinent concerns less amenable to scientific characterization.” In the climate change arena, for example, a focus on the expected costs and benefits of reducing carbon emissions has sometimes led to the disregard of low-probability catastrophic hazards and of extreme events that cannot yet be identified, let alone quantified.'* Rational policy making in this area should consider such extreme events, however, as well as concerns of intragenerational and intergenerational equity. Important factors that are often overlooked in the assessment of new technologies—in addition to uncertain and unknown hazards—include loss of economic security, compromise of traditional practices, and ethical matters. In other words, the “technical rationality” of risk assessment, which focuses on scientific measurement, may differ quite dramatically from the “cultural rationality” often reflected in the attitudes of nonscientists.!” Popular views on risk matters, informed by cultural rationality, consider quantifiable effects as well as contextual factors such as personal experience and social values. Thorough technical analysis is an important part of sound technology management, but incorporating social values and other concerns is equally critical. A more comprehensive assessment of technology would build on the exploration of these concerns to consider broadly the purposes of technology and the conditions we might place on its development and use.

Public Participation

In addition to analyzing technological consequences, technology assessment should engage the public in technology decisions. Participation by the broader public and incorporation of social values in technology management are essential because of the pervasive influence of technology. At a general level, technology shapes society and its institutions. And at an individual level, technology shapes people’s lives and affects their vital interests, including health, relationships, employment opportunities, and cultural practices. However, as science and technology policy scholar Daniel Sarewitz observes, “[T]he pursuit of technological transformation is largely exempted from formal democratic processes of eliciting value preferences and adjudicating value disputes about desired future states, even though technological innovation strongly expresses those very things.” Public participation in decisions regarding new technologies rests on several basic rationales: instrumental, normative, and substantive.?! The instrumental rationale for participation is a relatively narrow one of facilitating support for technological innovation and acceptance of new products. Instrumentally oriented participation treats the public as object rather than as partner in technology development and largely reinforces existing power structures.” The participation envisioned in this book, however, would have broader aims of advancing normative and substantive goals. Normatively, public participation in technology management is inherently valuable in that it reflects principles of democratic governance, social justice, and equality.”

Democratic ideals of autonomy and freedom from excessive government control call for citizens to have an active and meaningful role in the management of powerful technological forces.* Moreover, under contemporary democratic notions of public reason and discourse, policymakers should engage the broadest possible array of societal interests.” Such engagement seeks to empower citizens and thereby ensure that policy decisions having fundamental effects on society—including decisions regarding new technologies—are not reduced to the agenda of dominant institutions or a small group of elites. A further rationale for public participation is that it can contribute to substantively better outcomes. At a basic level, the public sometimes provides useful insights or suggestions that might otherwise be overlooked. More important, public input into technology management decisions can inform the policy-making process with public values and preferences, such as those regarding tolerance for risk and uncertainty, desire for change, and willingness to make trade-offs. Who is “the public” that should be involved in decisions on technology? Under one conception of the public, set forth by philosopher John Dewey, the public encompasses all persons substantially affected by the consequences of an activity.’ Several observations follow from this description. First, there may be multiple publics, depending on the activities that are at issue. Although different technologies may have different associated publics, emerging technologies increasingly have such broad and sweeping scope that their publics largely overlap. Indeed, since such technologies affect nearly all persons, the public impacted by a specific technology may not differ in any meaningful way from what we understand as “the general public.” Second, publics are or can be created. Emerging technologies often generate effects or potential effects on persons who are not yet aware of these effects. Genetically modified organisms, for example, are widely present in processed foods, yet few consumers are aware of this fact. In such instances, the affected persons may be shaped into an engaged public by raising awareness of technological developments and encouraging affected persons to apply their foundational values and beliefs to new circumstances. Third, given the impracticality in our complex society of creating a citizenry that is fully participatory on all issues of public significance, the idea of public participation necessarily implies representatives who speak on behalf of others having shared interests.” In addition to persons formally elected, these representatives may include those who lack formal power but nonetheless speak for others and not simply themselves.

Public input is warranted not only when explicit policy decisions are made regarding a technology affecting the public but also in the technical, expert-dominated process of technology assessment. ‘Technology assessment is riddled with value-based judgments that should be attuned to the values of the public and not merely the values of the experts conducting the assessment. Take, for example, the practice of risk assessment as applied to chemical substances. Although risk assessment is often characterized as a purely scientific, values-free undertaking, the values of those performing risk assessments necessarily influence the assumptions made, inferences drawn, and calculations performed.’* The subjective judgments involved in assessing chemical risks include judgments about what exposure levels are likely, which health consequences should be analyzed, how uncertainties should be interpreted, and how risk information should be framed.

Each of these judgments wrestles with problems of uncertainty. This is especially the case with respect to emerging technologies, where data on hazards are often sparse and where future applications may be unknown. Dealing with the uncertainty inherent in risk assessments and technology assessments is not a scientific issue; rather, it is fundamentally a policy issue that turns on subjective beliefs, values, and experiences.’ For this reason, incorporating public input and discourse into even the technical aspects of risk assessment and technology assessment is essential. Although lay citizens may have limited knowledge on technical matters and are sometimes misled by popular understandings of science and technology,” experience has shown that laypersons can communicate effectively with experts and contribute greater breadth and depth of knowledge and experience to technology assessments.*' Rational discourse involving citizens, experts, and other interested parties ultimately can lead to an expansion of relevant considerations and to the development of consensus regarding the common good.”

This emphasis on involving the public more directly and at earlier stages in technology management responds to a number of flawed assumptions of conventional technology assessment. First, the view that technology assessment was a purely objective process failed to take into account the subjective factors that inevitably enter into such assessments. Second, the assumption that technology assessments could focus on finished technologies overlooked the difficulty of controlling or modifying technologies that are often backed by substantial investments and powerful interests. Third, the conventional approach incorrectly assumed that technology policy decisions by elected officials would adequately reflect public values. In theory, political representatives are competent to make decisions consistent with the public interest and public values. Although this theory is foundational to our representative government,” there is a sound basis for concluding that elected officials inadequately represent the public with respect to technology matters.

Several factors systematically bias lawmakers against focusing on the potential risks of new technologies. These factors include the specialized body of knowledge necessary to understand science and technology issues as well as electoral pressures on politicians to respond to current headlines.** Few political rewards result from addressing issues having distant time horizons, such as emerging technologies. Not surprisingly, our government’s general policy toward uncertainties, such as those surrounding the health and environmental effects of emerging technologies, is one of willful ignorance rather than precautionary action.” In addition, as public choice theory predicts, lawmakers tend to cater to the interests of an organized and vocal minority at the expense of a majority whose interests are more diffuse.” This tendency is especially pronounced with respect to emerging technologies, which often have vigorous corporate or institutional advocates but little opposition from a public largely unaware of a technology or its potential risks.*’ All of these obstacles to the expression of public values through ordinary channels of representation warrant more direct and active public participation in emerging technology management.

In instances where the government does turn its attention to emerging technologies, the empirical techniques of risk assessment and risk-benefit analysis often dominate its policies to the detriment of public preferences.** ‘Technically, risk management (the values-driven, policy-making process of deciding how to respond to risk data) can be distinguished from risk assessment (the expert-driven process of identifying, analyzing, and quantifying risks).*? Notwithstanding this distinction, risk managers often defer to the quantitative data of risk assessment (or the lack of such data) in deciding how to respond to the hazards posed by new technologies.*” Because empirical techniques fail to account for all of the factors relevant to social decision making, however, an approach centered on quantifiable risks hides difficult, values-based choices under a veneer of objectivity. Indeed, the act of engaging in the discourse of risk through the risk management process reinforces the questionable assumption that risk can be understood, comprehensively measured, and effectively managed.“

In sum, technology assessment, once characterized as the exclusive province of experts, can and should be opened up to public participation. This broader and more open assessment should occur during and not just after technology development, thus bringing public concerns into research decisions and technology design. Care must be taken in the design of participatory processes to guard against the potential for simply reinforcing existing power relations or creating meaningless exercises in outreach. Nonetheless, if incorporated into properly designed processes, ongoing public participation can improve the quality of analysis and make the decision process more democratic.”

Our Experience with Technology Assessment

Past and present efforts in the United States to analyze and manage technology’s consequences fall into three basic categories: (1) formal technology assessment, which was practiced by the Office of Technology Assessment (OTA) until 1995 and has since been undertaken occasionally by other government agencies; (2) environmental impact assessment, which evaluates the potential health and environmental effects of policy choices and serves as a form of technology assessment when applied to technology policy decisions; and (3) nongovernmental assessment, which is carried out by technology developers themselves or by other nongovernmental entities. An examination of these efforts reveals that they fall short in advancing the analytical and participatory functions that technology assessment should achieve.

Formal Technology Assessment The Office of Technology Assessment

The concept of technology assessment arose as “critical voices within science began calling for preassessment before committing society to innovations such as supersonic transport and nuclear weapons.”* In the United States, Congress created the OTA in 1972 to provide nominally objective analyses that would inform policy decisions on technology matters. While the OTA earned high marks for the quality of its work on issues ranging from the feasibility of the Star Wars missile defense system to the cleanup of nuclear weapons laboratories, it ultimately played only a modest role in influencing U.S. handling of new technologies.**

Congress created the OTA at a time of increasing government spending on scientific research and development and growing social unease regarding the negative effects of science and technology. In the words of the statute establishing the OTA, modern technologies were “increasingly extensive, pervasive, and critical in their impact, beneficial and adverse, on the natural and social environment.”* Against this backdrop, Congress recognized the need for an independent source of technical information and objective analyses on which to base its decisions. The OTAs mission was “to provide early indications of the probable beneficial and adverse impacts of the applications of technology and to develop other coordinate information which may assist the Congress.”

Although the OTA originally was expected to perform long-term analyses of entire technological fields, in practice, technology assessment at the OTA was fairly constricted. Congress’s requests for information, which tended to emphasize specific subjects and require rapid responses, were typically met with short-term, narrow policy analyses.“ In an effort to maintain credibility and avoid alienating legislators who might disagree with its conclusions, the agency generally sidestepped recommendations and confined itself to reports on the technical components of the issues.** More significantly, technology assessment came to be understood not as a process but as a product: the reports generated by the OTA.” These developments were not surprising, given the OTA’ subordinate role and limited mandate. The OTA had no regulatory power and limited practical advisory power. The OTA ultimately never fulfilled its potential to study technologies in a broader and more systematic way or to transform how Congress and society related to technology.

The OTA eventually became unable to maintain the difficult balancing act of providing objective information to Congress while securing political support for its survival. Shortly after the 1994 Republican takeover, Congress eliminated the agency as part of the new majority’s promise to enact its “Contract with America.”*° Among the primary reasons given for eliminating the agency was its slow pace relative to congressional timetables. The OTA’ assessments, though limited in scope, often took one to two years to complete and thus often failed to meet the more pressing political needs of the members of Congress. Also contributing to the OTA’ elimination were accusations of bias—whether justified or not—commonly made by parties whose interests were undermined by the OTA’ reports.°!

America’s experiment with formal technology assessment through an agency dedicated to the task came to an end with the demise of the OTA. The needs for technology assessment and for deliberate management of technology, however, have not diminished. To the contrary, those needs are even greater today, as rapidly developing technologies will have widespread and long-lasting impacts and raise serious social and ethical concerns.

Institutional Alternatives to the OTA

In the OTA’ absence, the function of technology assessment has devolved to a number of other government institutions. Such assessments, however, have occurred on a narrow and ad hoc basis, if at all, and their scope and impact have been limited by traditional and legal constraints. The Government Accountability Office (GAO), for example, has conducted a few technology assessment pilot projects at Congress’s direction.” With the exception of a recent report on geoengineering, these analyses have focused mostly on narrow topics related to counterterrorism, such as the use of biometric technologies for border security and the use of cybersecurity measures to protect infrastructure. The credibility of the GAO’s work, moreover, is open to question, given the agency’s traditional expertise in performing audits, its limited experience with predictive assessments, and the past use of its reports for partisan objectives.” Another institution, the Office of Science and Technology Policy (OSTP), advises the president on science and technology issues. The OSTP has the authority to “initiate studies, including technology assessments, to resolve critical and emerging problems.”** With its fairly limited resources, however, the OSTP has focused primarily on serving as a channel of communication between the president and the scientific community and coordinating science and technology policy across the federal government.” Finally, the National Research Council (NRC) issues reports on science and technology topics in response to congressional and agency requests. Committees of leading experts prepare these authoritative reports through an extensive, peerreviewed study process. While well respected, these reports require significant time and resources to produce, and they are designed to generate expert recommendations in response to specific questions.’ The reports do not raise issues independently, address broader policy questions, or seek to foster public debate—activities that technology assessment should promote.”

NEPA as Technology Assessment

Technology assessment also can take place in the context of environmental impact assessment. In particular, the National Environmental Policy Act (NEPA) requires federal agencies to assess the environmental impacts of their actions.’ NEPA contemplates that this assessment be done in a public process, and its regulations provide for extensive public involvement in that process.” To the extent that federal agency actions involve the development or implementation of technology, NEPA analysis could serve as a form of technology assessment with respect to health and environmental effects. The federal government’s extensive role in sponsoring research and development, establishing technology policies, and possibly regulating technology’s adverse impacts underscores NEPA’ potential reach.® Nevertheless, narrow interpretations of NEPA by agencies and courts have minimized the statute’s value in technology assessment.

The Basics of NEPA

Sometimes described as a “Magna Carta” for the environment,*! NEPA was enacted with the purpose of integrating environmental values into national policies and planning processes. The central requirement of NEPA is the environmental impact statement (EIS), a document that must be prepared for “major Federal actions significantly affecting the quality of the human environment.” In the EIS, federal agencies must describe the environmental impacts of the proposed action, alternatives to the proposed action, and any irreversible and irretrievable commitments of resources that the proposed action would involve. As the Supreme Court has recognized, the EIS requirement is meant to ensure that the agency, in reaching its decision, will have available, and will carefully consider, detailed information concerning significant environmental impacts; it also guarantees that the relevant information will be made available to the larger audience that may also play a role in both the decisionmaking process and the implementation of that decision.”

For federal actions deemed to have environmental impacts that are less than significant, agencies still must prepare an environmental assessment (EA). An EA is a more limited inquiry aimed at identifying the environmental consequences of a proposed action and documenting an agency’s determination that those consequences are not significant.

NEPA was not meant merely to impose a requirement on agencies to document environmental impacts. Rather, NEPA was intended to achieve a wholesale reorientation of governmental actions and values, sensitizing agencies to environmental effects. Through preparation of a programmatic EIS, for example, an agency could evaluate environmental impacts of entire programs instead of conducting piecemeal assessments of individual projects. Overarching and comprehensive analyses presumably would motivate agencies to reorient programs in light of overall impacts.

Implementing and Interpreting NEPA

As implemented by federal agencies and interpreted by courts, however, NEPA has not fulfilled the goals embodied in its broad statutory language and legislative history. NEPA’ primary mandate, the requirement that federal agencies prepare EISs, has turned out to be less powerful than some of its crafters might have envisioned. First, that mandate applies only to “major federal actions”; it does not apply to private actions or to the majority of legislation enacted by Congress. Moreover, the Supreme Court has construed the EIS requirement as imposing only a procedural duty rather than as mandating substantive results.” This does not necessarily make the EIS requirement toothless: Federal agencies, having analyzed the environmental ramifications of a proposed action and considered public concerns, may make more environmentally sound decisions than would otherwise be the case. Nonetheless, agencies sometimes view the EJS as little more than a paper-pushing hurdle to be overcome before the agency proceeds with a predetermined course of action.

Supreme Court decisions have played a critical role in constraining the statute’s reach and effectiveness. The Court first addressed NEPA in 1975 in Aberdeen & Rockfish R.R. v. Students Challenging Regulatory Agency Procedures (SCRAP). There, the Court rejected the argument that an EIS must be integrated into the earliest stages of an agency’s decision-making process. Rather, an EIS must be prepared only when an agency “makes a recommendation or report on a proposal for federal action.” The Court subsequently held in Kleppe v. Sierra Club that neither the contemplation of a project nor the preparation of studies of possible development triggers the obligation to prepare an EIS absent a proposal for federal action.” Together, SCRAP and Kleppe allowed agencies to defer analyzing environmental impacts until after the overall policy-making process is well underway. These decisions did not specifically involve new technologies, but they offered little hope to those expecting that NEPA would compel agencies to conduct effective technology assessments.

The Supreme Court made its most direct pronouncements regarding the analysis of new technologies under NEPA in Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc.’' In this challenge to licenses granted to specific nuclear facilities, the Court rejected the notion that NEPA provides a forum for the wholesale consideration of the desirability of new technologies: