Abstract and Keywords
This book offers a variety of efforts to bring into view changes along two axes—changing ways of governing risk and changing ways of doing the sociology of insurance and risk. They represent the beginning of what would be an important new area of thought and research. This chapter considers each axis in a bit more detail. Although the embracing of risk represents a systematic change in thinking about institutions, with the potential to transform the social contract, it is important to emphasize two kinds of limitations. One is the endurance of the risk-spreading infrastructure. The second limitation relates to risk taking.
Centuries are too convenient not to be used for organizing historical developments. But as cognitive psychologists have discovered, people are so hard-wired for seeing patterns that we tend to see them where there are none. For this reason, historical patterns witnessed at the start of a new millennium may be sufficiently suspect members of an already highly suspicious class that they ought, perhaps, to be discounted entirely. With that said, however, a premise of this book is that we are, at the beginning of the twenty-first century, witnessing an important transformation in the approach to risk and responsibility, an approach that we call “embracing risk.”
In coining the term embracing risk, we mean to both evoke and distinguish the idea of “spreading risk” that has been so influential over the last century. Embracing risk captures two related cultural trends. The first follows logically from efforts to spread risk and consists of a wide variety of efforts to conceive and address social problems in terms of risk. For example, money management, social services, policing, environmental policy, tort law, national defense, and a host of otherwise unrelated fields have all come to share a common vocabulary of risk. The second cultural trend is a reaction against spreading risk, and it consists of various efforts to make people more individually accountable for risk.
Together these two trends mean that, as more of life is understood in terms of risk, taking risks increasingly becomes what one does with risk. Once you begin to look, you can see these efforts almost everywhere: in transformations in the private insurance market, pensions, and social insurance; and in popular culture, where extreme sports, day trading, and the (p.2) cult of the technology entrepreneur all exalt individuals who (at least seem to) spurn the safety nets of large institutions.
Ideas like embracing risk or spreading risk inevitably promise more than they can deliver, and they never capture more than a partial vision of a cultural moment. Yet that does not diminish their importance. The notion of insurance as risk spreading was a veritable genie from a bottle in the early twentieth century, as reformers sought to extend its logic from workplace accidents to automobile accidents, unemployment, poverty, disease, and nearly every other social problem, with wide-ranging consequences (Simon 1998). We see signs that embracing risk may become as symbolically powerful. Indeed, reformers today are recasting their approaches to many of those same problems in terms of embracing risk.
This embrace of risk creates new challenges, not only for those with risks to spread or embrace, but also for the study of risk and insurance. Historians, anthropologists, political scientists, and lawyers have all explored the emergence of modern approaches to risk and insurance. But the implicit background for much of this work has been the risk-spreading approach. For example, this work has almost always understood insurance as a mechanism of risk spreading, and risk itself in terms of future harms with measurable probabilities (Ewald 1986).
The two trends that come together in embracing risk challenge us to understand other aspects of both risk and insurance. Insurance institutions that embrace risk push us to recognize that insurance can be about much more than risk spreading. Looking back with this expanded vision, we see that insurance has always done much more than spread losses. Similarly, if risk is something to be embraced, it cannot only be about harm or danger. And, once we free risk from risk spreading, there is no longer a reason to confine risk to probability. Once again, looking back with expanded vision we see that risk was never completely tied up with either harm or probability.
The essays in this book are a variety of efforts to bring into view changes along two axes—changing ways of governing risk and changing ways of doing the sociology of insurance and risk. They represent the beginning of what we hope will be an important new area of thought and research. We have collected these essays together to invite and encourage further work along either of the above axes or between them. Toward that end, in the remainder of this introduction we consider each axis in a bit more detail. Introductory sections at the beginning of each of the two parts of the book describe the individual essays that follow them and how each essay contributes to the book's main themes.
(p.3) From Spreading Risk to Embracing Risk
From the adoption of workers compensation laws in the early twentieth century through at least the late 1980s, the United States and other industrializing societies socialized, or spread, more and more risks. Over this period, ever-expanding public and private insurance pools assumed financial responsibility for significant risks faced by individuals, families, and organizations. On the private side, the twentieth century witnessed the dramatic growth of health insurance, tort liability insurance, workers compensation insurance, and private pensions (which typically have an annuity, and, hence, insurance character), as well as the slower but still steady rise in older forms of insurance such as life, property, and disability insurance. On the public side, there was the creation and perhaps even more dramatic expansion of an entirely new social insurance sector, beginning with the multifaceted Social Security program during the New Deal and followed by Medicare, Medicaid, and natural disaster insurance, as well as a host of public sector insurance ventures directed primarily at business risks.
During this era, insurance was widely understood as the science and art of spreading risks over populations. Indeed, some visionaries went so far as to claim that insurance embodied the superiority of science and technology over religion. Through insurance, science would be able to resolve social conflicts and produce forms of collective mutuality envisioned but unachieved by any of the major religions (see, e.g., Dawson 1895).
The trend was not always uniform, and significant risks and important segments of the population were always left out. Nevertheless, the overall picture during this period is one in which the steadily employed, as well as an expanding percentage of those outside the labor market, enjoyed increasing protection from the financial consequences of illness, injury, old age, premature death of the family breadwinner, fire, and natural disaster. Indeed, “more insurance for more people” might best describe the twentieth-century U.S. domestic social policy well into the Reagan/Bush years. Our focus is on the United States, but we think that experience is not unique. If anything, risk has been socialized to an even greater extent in Western Europe and Japan.1
More recently, there has been a series of developments in the United States and elsewhere that suggest the appeal of a domestic social policy of “more insurance for more people” has begun to fade in favor of policies that embrace risk as an incentive that can reduce individual claims on collective (p.4) resources. Significantly, these developments are occurring in both public and private forms of insurance, so that they cannot be attributed solely to a reexamination of the role of the state in the distribution of risk and responsibility.
Above all, these developments reflect an increased concern about how people react to being protected from risk and a new emphasis on the need to manage incentives to curtail what is perceived as the runaway growth of public and private insurance programs. This focus on what Carol Heimer (1985) has called the “reactive” nature of risk has led to outright reductions in benefits in some social insurance programs, such as workers compensation and public welfare programs. It has also led to the increasing coordination of the finance and delivery of insured services, manifested most obviously in the United States by the shift to managed health care in both private and public health insurance, but also by related developments in property and liability insurance and workers compensation. Across a wide range of institutions, officials are now as concerned about the perverse effects of efforts at risk shifting as they are about the risks being shifted. For example, rather than seeking to eliminate poverty (a master goal of the “more insurance” era), current social policy aims at eliminating welfare.
While these changes are often seen as a shift against the poor in favor of the well off, and against consumers and in favor of business, they affect the middle class as well as the poor, and businesses as well as individuals. For example, private pensions, annuities, and life insurance are engaged in an historic shift of investment risk from broad pools (the classic structure of risk spreading through insurance) to individual (middle-class) consumers and employees in return for the possibility of greater return. Many annuities and pensions have moved from a “defined benefit” approach, in which the risk of sufficient revenues to pay the promised benefit is on the collective pool, to a “defined contribution” approach, in which the risk of sufficient revenues for a comfortable retirement remains on the individual. (Remarkably, this significant change in the circumstances facing an increasingly large proportion of U.S. families is taking place with almost no public discussion.) The claim is that embracing risk will provide workers with greater returns with which to enjoy their retirement.
Similar developments are also occurring in insurance products sold to large businesses. Since the 1980s there has been significant growth in alternative risk mechanisms addressing property and casualty risks. Many of these reward businesses for retaining risk. Examples include captive insurance companies, fronting, third-party administrators, and finite risk insurance; in each of these, the individual business retains most of the risk while (p.5) obtaining the tax and risk management benefits of the insurance form. Similarly, employers increasingly are self-insuring for workers compensation and health insurance risks, and there is a related trend toward larger deductibles, retroactive premiums, and other similar ways of retaining risk by organizations that continue to use traditional insurance. In addition, although this is a controversial claim, there may be an increasing use of corporate structure to insulate assets from risky activities, with a corresponding reduction in the ability of the tort/liability insurance system to socialize risk (Lopucki 1996).
In the public arena, the most significant manifestation of this trend may have been the discussion of social security reform in the 2000 presidential campaign. Historically, the Social Security system has been the most visible example of the “spreading risk” approach. The government collects a fixed percentage of earned income (up to a salary cap) through a special payroll tax that functions as a social insurance premium. In return the government provides a guaranteed annual income beginning at retirement or disability, keyed to a person's earnings history and intended to keep the elderly and disabled above poverty. This system spreads the risk of physical inability to work among almost the entire working population.
The Republican candidate, George W. Bush, and the Democratic candidate, Al Gore, both proposed creating a market-linked segment of the Social Security retirement benefit that represented a significant move away from the spreading risk approach. Both plans would have placed on individuals the risk that market forces will result in disappointing or even zero returns. Although Bush proposed a more dramatic departure from tradition (under which participants could place as much as 2 percent of their income into an individualized private investment fund), the details of the respective plans are less important here than the fact that both represented a new emphasis on risk and reward.
Popular culture and consumer behavior also reflect the increasing attraction of risk-taking activities. Perhaps most significant is the dramatic expansion of stock market investing, particularly in its “extreme” version—day trading. Long considered of little value or interest to middle-class, wage-earning Americans, the stock market since the 1980s increasingly has become an object of intense mass appeal. By the end of the twentieth century, participation in the market reached extraordinary levels, and all manner of supporting activities proliferated as well—web sites, magazines, television and radio shows.
This broad focus on investing is often attributed to the bull market of the 1990s, and no doubt the repeated stories of success fueled the popular (p.6) growth in investing (much as news of gold strikes in California helped set off that “rush”). But the appeal of investing was also enhanced because of its resonance with the larger cultural trend of embracing risk. This was particularly evident in day trading, in which ordinary individuals give up their day jobs to become full-time market players using the web and discount brokers. Here the objective evidence suggests that, even at the height of the market, most people lost money, and many lost badly. Thus the popular interest in day trading seems to us to reflect less the promise of easy wealth than the cultural attraction of embracing risk. It represents the possibility of attaining autonomy, leaving behind the frustrations of working for someone else, by risking your own capital.
Beyond the stock market, numerous other activities promising the possibility of wealth or thrills, or both, have moved from the illegal and tawdry to the legal and acceptable, including gambling (once illegal, now state sponsored in most parts of the United States), extreme sports, and adventure travel. As with the stock market, the influence of direct participation expands many times when one considers those caught up in reading, watching, listening, and thinking about all these ways of embracing risk.
Although the embrace of risk represents a systematic change in thinking about institutions, with the potential to transform the social contract, it is important to emphasize two kinds of limitations. One is the endurance of the risk-spreading infrastructure. As François Ewald notes in the final chapter of this volume, insurance institutions remain central even as the coherence and appeal of what he calls the “solidarity” paradigm (our risk spreading) decline. Just as individual fault and responsibility remained part of modern legal culture even with the growth of risk spreading and social insurance in the early twentieth century, so too will risk spreading survive the embrace of risk. For example, the middle classes of the United States have shown considerable resistance when faced with immediate reductions in their own safety nets, despite their enthusiasm for programs aimed at saving the poor from dependence on others. Indeed, we see the embrace of risk more as a recognition of the limits of risk spreading than as a wholesale replacement for it.
The second limitation relates to risk taking. While some kinds of risk taking have reached an expanded class base (such as investing) and others a newly upgraded cultural status (such as gambling and extreme sports), others have become subject to even greater anxiety and harsh repression (such as economic immigration, drug taking, and sexual experimentation). Moreover, as François Ewald suggests in his chapter, some kinds of risk taking accepted in the era of risk spreading, ranging from consumption of new (p.7) technologies to the likelihood of casualties in military service, have become subject to inconsistent but often powerful public resistance and demands for “zero risk.”2 Clearly, taking risks is only one part of a complex emerging configuration of risk that also includes new demands for precaution and even abstinence. Significantly, however, “taking risks,” “zero risk,” and “just say no” all reject the more utopian claims of the risk-spreading evangelists in favor of a new responsibility for the risks imposed on oneself and others.
Risk and Insurance
Insurance is the paradigmatic risk-spreading institution, and most discussions of insurance focus, appropriately, on its risk-spreading features. Not surprisingly given our focus, risk spreading occupies a comparatively small place in the extensive discussion of insurance in this volume. Nevertheless, we would not want anyone to think that we ignore the risk-spreading aspects of insurance. After all, the growth of the “insurance state” was predicated on the ability of social insurance institutions to spread losses (Ewald 1986). Nevertheless, there are other, less often considered aspects of insurance that the embracing risk approach leads us to emphasize.
Because almost everyone, at least in the developed world, has so much experience with insurance—buying auto, life, and homeowners insurance; consuming medical services covered by health insurance; paying social insurance taxes; and so on—insurance is a category that people tend to believe they understand. Yet, as anyone who has thought deeply about insurance can report, identifying its core features for the purpose of clearly distinguishing it from what is not insurance is anything but a simple task. Spencer Kimball, one of the leading insurance law scholars of the twentieth century, began a textbook he wrote late in his career by writing: “There is no good definition of ‘insurance,’ for any purpose. This book will not seek to provide one” (Kimball 1992: xxv).
Following the French social theorist (and our contributor) François Ewald, though employing a somewhat different framework, we distinguish among four aspects of insurance: institutions, forms, technologies, and visions (Ewald 1991). We find these categories useful because they demonstrate the conceptual variety of the activities lumped under the label “insurance.” For this reason, the precise definitions of the categories are less important than the sense of the multiple aspects of insurance that they capture.
The first two of these categories, insurance institutions and insurance (p.8) forms, have intuitively obvious meanings. Insurance institutions are the various kinds of organizations that provide insurance. Federal social insurance agencies (such as the U.S. Social Security Administration) and stock insurance companies (such as Allstate Insurance Company) are two (very different) insurance institutions. Insurance forms are the various kinds of insurance provided by insurance institutions, as well as the variations in form among those kinds. Life and property insurance are two different forms of insurance. “Whole” life insurance (life insurance that builds a cash value and, thus, incorporates a savings element) and “term” life insurance (life insurance that does not include a savings element) are two different forms of life insurance. Fire insurance and flood insurance are two different forms of property insurance.
Insurance technology is the “how to” of insurance. Examples include the mortality tables, underwriting classifications, and inspection procedures of ordinary life insurance; the incentive-based medical provider contracts, retrospective review, and computerized claim processing procedures of managed health care companies; the payroll tax, disability schedules, and administrative review procedures of the social security program; and the standard-form insurance contracts used in almost all private insurance.
Insurance in this sense of insurance technology refers to a set of procedures for dealing with risk.3 The concept of self-insurance may help illustrate what it means to understand insurance as a technology. An organization that self-insures with respect to a particular risk does not purchase insurance for that risk. Thus, at the risk of belaboring the obvious, that organization has no insurance for that risk. Yet self-insurance differs from no insurance in that an organization that self-insures does not simply forgo buying insurance, it also adopts (at least in theory) a set of self-insurance procedures. These procedures are similar to those an insurance company would apply—for example, keeping track of past losses and setting aside financial reserves to meet comparable losses in the future.
As a technology for managing risk, insurance extends far beyond what might ordinarily be understood as the insurance field. Insurance institutions have developed many ways of managing risk that other people and institutions have adopted. The great life insurance companies were pioneers in epidemiology and public health. The fire insurance industry formed Underwriters Laboratories, which tests and certifies the safety of household appliances and other electrical equipment. Insurance companies seeking to cut their fire losses formed the first fire departments. More recently, health insurance companies have been behind many efforts to compare, test, and measure the effectiveness of medical procedures. Focusing on insurance (p.9) technology helps us to see that insurance not only spreads risk, it also does all the other things that these insurance technologies do.
Finally, insurance visions are ideas about and images of (or, alternatively, discursive practices regarding) insurance that animate the development of insurance technologies, institutions, and forms.4 Insurance visions are the aspects of insurance that we are most interested in, yet they are less easily described than insurance technologies, institutions, and forms. The best way we know of explaining what we mean is by example.
The first example comes from work by the sociologist Viviana Zelizer (1979). As Zelizer has described, it was once commonly believed that life insurance was immoral, either because it represented a presumptuous interference with divine providence, because it was seen as a form of gambling, or because it impermissibly equated life and money. This vision of insurance has had important consequences for the development life insurance institutions in the West. It slowed the growth of life insurance in at least France and the United States, and it helps explain the intense preoccupation in the United States during the mid-nineteenth century with establishing the morality of all kinds of insurance (Zelizer 1979; Baker 1996). As discussed in chapter 2, related ideas about risk and insurance more generally affect Islamic institutions today.
A second example comes from earlier work by Tom Baker (1994) addressing the role of insurance visions in insurance contract law. Because the primary benefit of insurance is a sense of security that for most people is never tested by a catastrophic loss, the value of insurance rests, in an important sense, in the imagination. Courts have recognized the importance of imagination to insurance by placing great emphasis on the “reasonable expectation of the insured” and by holding that insurance advertising is relevant evidence for determining that expectation. In that advertising, insurance companies evoke a vision of insurance that differs from the vision of insurance they employ when denying claims. (See also Stone 1994.) Their “sales” vision is the promise “to be there,” and it is dominated by narratives of family and the need to protect the individual against sudden misfortune. Their “claims” vision is a complicated amalgam of tough love and protecting the insurance fund, and it is dominated by narratives of institutional ethics and the need to protect ratepayers against fraud and abuse. As described by Baker, courts first decide which of these visions to employ in resolving insurance contract disputes. Who wins a dispute often depends as much on which vision the court adopts as on how the court applies that vision.
A third example is the actuarial vision of insurance that we and others have described (Simon 1988; Baker 2000; Pal 1986). In the actuarial vision, (p.10) the ideal type of insurance involves premiums paid in advance, guaranteed indemnity in the event of a covered loss, and risk-based premiums based on the best available information regarding the expected losses of the individuals insured. This vision of insurance has had enormous consequences. It helps explain the decline of fraternal insurance in Britain and the Unites States over the nineteenth and early twentieth centuries, as actuarial expectations overcame the values of “friendship, brotherly-love and charity” (Doran 1994). It helps explain the decision to model unemployment insurance on private insurance, and the related effort to tightly link benefits to premiums (Pal 1986). And it helps to explain the intensity of the popular belief that Social Security retirement benefits have been earned by the people who collect them, as well as the corresponding expert belief that Social Security is not really insurance (because the money to pay today's retirees comes from the contributions made by today's workers and not from the contributions of the retirees themselves). Indeed, the actuarial vision of insurance has been so successful that many well-informed people would deny that it is a vision at all and assert, instead, that it is the model of insurance.
Finally, conceptions about risk in general, and the embrace of risk in particular, are also examples of insurance visions. The idea that some amount of risk is good for people and that too much protection is harmful has important consequences for the development of insurance technologies, institutions, and forms, as well as the development of personal identity and social institutions other than insurance. Many of the essays in this book explore these consequences.
As we hope this brief discussion illustrates, this framework of insurance institution, form, technology, and vision opens a more expansive view on the workings of insurance than the relatively simple (but very important) concept of risk spreading. We will not in this book ignore risk spreading. The fact that insurance spreads some losses and not others is an essential feature of the distributional concerns addressed in many of the chapters. But even there it is not the risk spreading per se that we find notable. Rather it is the visions of insurance that animate decisions about whose and which losses to spread.
Risk(s) Beyond Insurance
The essays in this book also address risk(s) beyond insurance in three senses. First, they consider the use of insurance technologies and visions to govern risk outside of insurance institutions. Second, they address the role of other kinds of institutions and practices in developing technologies and visions for dealing with risk (some of which in turn may influence (p.11) insurance). And, third, they address risks that are beyond insurance in the sense that they cannot or, perhaps should not, be governed by insurance. The first two we think of as addressing risk beyond insurance, and the third as addressing risks beyond insurance. Hence the heading “risk(s) beyond insurance.”
From our opening description, it might seem that the vision of embracing risk refers only to the latter idea: that, for the benefit of individual or social welfare, some risks cannot or should not be transferred to insurance institutions. Yet we think embracing risk also encompasses a broader idea of governing through risk. Though it can hardly be summed up in a single sentence, the core idea of governing through risk is the use of formal considerations about risk to direct organizational strategy and resources. Indeed, for those who equate risk with insurance (Ewald 1991), this is one way that insurance produces a change in the culture.
There are many contemporary examples of governing through risk. Money managers develop portfolios at the risk–reward frontier (Bernstein 1992, 1996). Social service agencies target at-risk children (see, e.g., Health and Human Services 1994). Community policing efforts are targeted at high-risk areas (Ericson and Haggerty 1997). Environmental engineers conduct risk assessments of hazardous waste sites and other sources of environmental concern (Graham and Weiner 1995). Mountain climbers rate peaks and climbs according to risk and climbers according to the risks they are qualified to take (Simon, chapter 8 this volume). Fraternities redefine gender relations in response to the risk of sexual harassment (Simon 1994). Judges and law reformers debate accident law in terms of the allocation and spreading of risk (Calabresi 1970). And presidents and prime ministers alike now define themselves by the risks they would have the state lift from the shoulders of the populations (criminal victimization, poverty in old age) and the risks they would not (poverty in childhood, economic security, health care) (Giddens 1999).
Admittedly, thinking of risk or risks as beyond insurance reflects a very insurance-centered view of the world. The world is full of risks and only some are handled by insurance. Why isn't the more appropriate question, “Which few of the many risks in the world are handled by insurance and why?” Surely that is a more coherent, manageable set of risks to identify and discuss. After all, insurable risks are by definition susceptible to the insurance technologies, forms, and institutions that the study of risk and insurance aims to explore.
Yet thinking about risks as beyond insurance does add something to understanding. Insurance institutions and technologies gather risks, concentrate (p.12) and contain (“pool”) them, shape and package them for resale on the secondary market, or, alternatively, plug them directly into the distributional circuits of the state. In the process, they spread and thereby, in an important sense, eliminate (or at the very least reduce the importance of) those risks from the day-to-day concern of people who are exposed to them. Risks that are beyond insurance are not spread and eliminated in this way, with consequences for the people exposed. A risk beyond insurance is experienced in a very different way than a risk that is insured.
In a society in which “more insurance for more people” has been so strongly encouraged, thinking of a risk as beyond insurance leads to an obvious question: Why? What distinguishes this risk from the other risks turned over to insurance? Is it uninsurable in some technical or practical sense and, if so, according to what logic and compared to what other risks that are insurable? Or, is it uninsurable in a normative sense—meaning that it should not be insured? If so, according to what logic, for whose benefit and, once again, compared to what other risks that are insurable?
Because of the crucial role of insurance institutions in socializing risk and responsibility (a role that Tom Baker explores in chapter 2), studying risks beyond insurance opens a window on the limits of social responsibility and the role of ideas about individual responsibility in the shaping of insurance institutions and forms. For example, when Jonathan Simon describes in chapter 8 the competition between the climbing ethic of summiteering and the climbing ethic of mountaineering, he is quite consciously developing a metaphor for application in the flatlands below.
Toward a Sociology of Insurance and Risk
Although there is a significant research literature on insurance and risk, it remains relatively small compared to the importance of insurance to society. Given the enormous size of insurance institutions relative to contemporary Western economies, it is surprising that most social scientists and historians have paid so little attention to insurance. Indeed, looking at twentieth-century governance, it is tempting to see insurance as the sleeping giant of power.
Consider just the following two examples of insurance setting social standards in the United States. First, as Carol Heimer discusses in chapter 6, insurance acts as a gatekeeper to homeownership. Homeownership is rightly taken to be one of the constitutive features of the American economy, with all its attendant influences on both consumption (more of it) and political preferences (more conservative). Not surprisingly, the availability of mortgages at an affordable rate is widely regarded as an important economic (p.13) indicator. The role of homeowner's insurance as a vital ingredient in the availability of mortgages is much less widely recognized, however. Obtaining homeowners' insurance is mandatory for standard mortgages, and homeowners' insurers employ their own underwriting concerns that are independent from the screening procedures implemented by the lender (Squires 1997).
Second, private insurance can be a crucial form of delegated state power. Rather than set its own criteria for access to vital economic freedoms like operating an automobile or a business (which would be politically controversial and even, perhaps, unconstitutional), the state mandates that a person wishing to engage in any such activity first obtain some form of insurance. Examples include liability insurance for automobile owners, workers compensation insurance for employers, and surety bonds for companies engaged in business with the state. In most cases, the state avoids providing the insurance and thereby asks the private market—typically property-casualty insurance companies—to set the underwriting criteria that will determine access to these privileges and immunities. Motivated by controlling losses they have contracted to pay, the companies set up their own norms of conduct, which they enforce by contract terms and pricing (and, ultimately, through the state judicial system).
Whether obtained as a result of compulsion or simple prudence, insurance is a form of regulation. Specific exclusions and conditions written into coverage for property, life, and health amount to a form of private legislation as binding as any the state legislature enacts. Significantly, this “legislation” acts inside the home or business, where the sovereignty of the king was traditionally expected to stop (O̓Malley 1991). Indeed, within a regime of liberal governance, insurance is one of the greatest sources of regulatory authority over private life.
Despite this central role, insurance has been almost completely ignored by the traditional humanities and social sciences, at least outside of economics departments and business schools. As a result, neoclassical economics is now the dominant paradigm for the analysis of insurance and risk. The leading insurance journals and academic departments (at least in the United States) are populated by economists. Policy debates over the nature and extent of public insurance and the regulation of private insurance are almost always framed in economic terms. Yet, although economic analysis has produced significant, insightful work on insurance (Baker 1996), that work has largely ignored institutions, history, and culture—precisely what we would emphasize.
There has been significant sociological work on some insurance institutions, (p.14) most significantly social insurance and, more recently, private health insurance. With some important exceptions in the area of economic sociology, however, most of this research operates at different levels and with different purposes than we envision for a sociology of insurance and risk. The work on social insurance roughly breaks down into three groups: work on the origins and political trajectory of the welfare state (see, e.g., Nonet 1969; Skocpol 1992); work on the interaction of social insurance organizations with their clients (see, e.g., Lipsky 1980); and work on the limits of social insurance in achieving redistributive ends (see, e.g., Piven and Cloward 1971). None of this work focuses broadly on insurance. Similarly, nearly all the work on private health insurance is focused exclusively on the effect of insurance on medicine and health (see, e.g., Starr 1982) and, thus, falls within the sociology of medicine rather than the sociology of insurance and risk.
Governmentality and Economic Sociology
The immediate intellectual roots for this project lie in two traditions: economic sociology and the “governmentality” literature stimulated by the later work of Michel Foucault. Because both of these traditions are relatively unfamiliar ones, and few people have read broadly in both, it is worth spending some time reviewing them. In the process, we describe how they contribute to the development of a sociology of risk and insurance. Our goal is not to set out a definitive statement or program of research, but rather to provide a sense of an emerging field of inquiry.
Economic sociology is a branch of sociology that shares many of the goals and concerns, and some of the methods, of institutional economics. In briefest terms, it involves the application of empirical and conceptual sociological methods to economic phenomena, and it dates back to the giants of classical sociology: Marx, Weber, and Durkheim (Smelser and Swedborg 1994). Economic sociology has largely neglected insurance and risk as demonstrated, for example, by The Handbook of Economic Sociology (Smelser and Swedborg 1994), which contains no significant discussion of insurance and only a brief discussion of risk. Nevertheless, the framework and methods of that tradition can be readily applied to insurance and risk. Indeed, Viviana Zelizer's (1979, (1985) and Carol Heimer's (1985) work on insurance and risk are directly within that tradition. Both played an important role in promoting the work that appears in this volume.
Zelizer's study of nineteenth-century life insurance documented the role of religious and other cultural ideas and practices in the development of the U.S. life-insurance market. In her first book, Morals and Markets, she (p.15) explained the dramatic expansion of the life-insurance market in the mid-nineteenth century as the result of ideological work by insurance entrepreneurs. The entrepreneurs successfully overcame profound moral and religious objections to insurance (objections that delayed the growth of the French life-insurance market even longer) by inventing and then promoting a vision of life insurance as a form of moral responsibility. In a chapter in her second book, Pricing the Priceless Child, she charted the growth of the market for life insurance for children as a cultural battleground. On the one side were insurance companies offering the poor security from the shame and fear of burying a child in a pauper's grave. On the other side were middle-class reformers who saw child insurance as an invitation to murder. By demonstrating the importance of culture to the development of the market, Zelizer helped reinvigorate economic sociology and began an important conversation with economists that the essays in this book hope to continue. By focusing on the role of culture in the development of insurance (and the role of insurance institutions in changing culture), she put insurance on the map of economic sociology.
Heimer's study of insurance contracting, Reactive Risk and Rational Action, also engaged a topic of great interest to economists: the moral hazard of insurance. A term with a very rich history (Baker 1996, 2000), “moral hazard” refers to the effect of insurance on incentives. If not properly managed, insurance can reduce the incentive to be careful to avoid a loss and it can reduce the incentive to manage the costs of recovering from loss. As Heimer put it, risk is reactive, not static. Working within a rational choice framework that she shares with economics, she explored empirically how insurance companies manage reactive risk through the use of insurance contracts.
Heimer's research on the contracting behavior of insurance companies provided the first detailed sociological study of private insurance as a form of social control. As her research revealed, control is the fulcrum for managing moral hazard. Losses over which the insured has no control do not present a moral hazard problem, because the reaction of the insured to being freed from risk does not affect the odds of loss. The less control the insured has over loss, the more willing insurance companies are to sell insurance, and the more complete that insurance will be. Where the insured has substantial control, however, the price for buying insurance often includes giving up a measure of that control. Insurance companies may demand that the insured institute safety procedures, undergo periodic inspections, or allow the insurance company to control the efforts taken to recover from loss. Examples include sprinkler requirements in commercial fire insurance (p.16) contracts, inspection clauses in workers compensation insurance contracts, and hospitalization precertification clauses in health insurance contracts. When insurance companies manage moral hazard, they are regulating behavior, not simply spreading risk.
Along with demonstrating that insurance is a form of regulation, Heimer's research also provided a historical, conceptual, and empirical framework for understanding the embrace of risk. The claim that embracing risk is good for individuals and society rests on the assertion that, in her terms, the reaction to being freed from risk can be as harmful as the insecurity that insurance aims to prevent. Heimer's study explored in a detailed, convincing fashion the circumstances in which insurance companies have required their policyholders to retain some risk, which creates what Heimer called a “community of fate” between the insurance company and the policyholder. This community of fate lessens the policyholder's incentive to react to insurance by reducing care. By documenting the extent to which many different kinds of insurance contracts create communities of fate, Heimer's research helps us see that insurance companies have always required people to embrace some risks.
Research inspired by Foucault's late work on governmentality also provides an important foundation for our project. The word governmentality is his neologism for “governmental rationality.” For Foucault, government is a much broader concept than the government of a political system. Indeed, rather than a noun referring to a set of institutions, he understood the word government to refer to a set of practices that are engaged in by individuals and institutions at every level of society. Think of the governor on a motor vehicle—a device that prevents the vehicle from going faster than a set speed—or the self-government of one's own behavior. “Governmental rationality” is “a way or system of thinking about the nature of the practice of government (who can govern; what governing is; what or who is governed), capable of making some form of that activity thinkable and practicable both to its practitioners and to those upon whom it was practiced” (Gordon 1991:3). The clearest illustration of the concept of governmental rationality in this book appears in chapter 5.
Our earlier description of insurance as a form of regulation already suggests the appeal of the governmentality literature to the study of insurance and risk. That literature provides a set of concepts and contexts for investigating governmental practices across a wide range of social fields and institutions. Within the governmentality literature, insurance is simply one of many forms of knowledge and practices that operate in the space between the individual and the state.
(p.17) In his lectures on governmentality, Foucault called attention to the gap between studies of power at the level of the individual (as in his History of Sexuality) and studies of power at the level of the state (Gordon 1991). The governmentality literature has begun to fill that gap. Our contributors François Ewald and Nikolas Rose have been perhaps the most important French and English language contributors to this line of research generally.
Ewald's framework of insurance technologies, institutions, forms, and visions is directly in this tradition, helping to explain how insurance functions as a form of government. His 1986 book, L̓Etat Providence traced the history of the welfare state by focusing on the emergence of risk as a dominant way of knowing and intervening in the world beginning in the nineteenth century. For Ewald, risk is the product of insurance technologies that bring probabilistic methods to bear on aggregated data, producing actuarial representations of risk as an object that can be known and distributed. Ewald's paradigm case was workers compensation, which aimed to rationalize industrial accidents by preventing harm where possible and then spreading the costs of accidents that could not be prevented. (See also Ewald 1991; Burchell, Gordon, and Miller 1991; Rose 1999.) Put in terms of our project, Ewald offered a genealogy of the paradigm of “more insurance for more people.”
Recent work by Rose (1999) has explored the breakdown of this social logic of governance and the emergence of other approaches that emphasize the individual. (See also Stenson and Watt 1999; Simon 1999.) Rose describes these approaches as “advanced liberalism,” linking them to the tradition of liberal thought that arose in the nineteenth century. The cultural trend we identify as embracing risk fits closely with the advanced liberal emphasis on the individual. Many of the policies Rose links to advanced liberalism seek to place more risk on the individual and to dismantle the large risk pools that socialized risk under the mandate of more insurance for more people.
Embracing risk is a strategy of government—a governmental rationality—that applies to many levels of society. As Jonathan Simon explores in chapter 8, embracing risk can be a strategy of individual self-government. Pat O̓Malley and Martha McCluskey find in chapters 5 and 7, respectively, that embracing risk can be a strategy of government by or through the state. And as Tom Baker, Carol Heimer, Martha McCluskey, and Nikolas Rose explore respectively in chapters 2, 6, 7, and 9, embracing risk can be a strategy of government by such intermediate social institutions as insurance companies, employers, banks, and health care institutions.
(p.18) The Risk Literature
The subject of risk has received more widespread scholarly attention than insurance. Yet most of the research on risk focuses on the needs and opportunities of existing insurance institutions, rather than on the social and cultural role of risk and insurance more broadly conceived.5 Notwithstanding that common orientation, risk studies is far from a unified field. Indeed, there are at least as many approaches to studying risk as there are academic disciplines (Krimsky and Golding 1992; Burger 1993). As a result, reading the risk literature across disciplines reveals almost as much about those disciplines as it reveals about risk. In that sense, the risk literature is as much a window on academic disciplines as risk management is a window on organizations, and as visions of insurance (and risk) are windows on social change.
At the risk of oversimplification, we lump much of the risk literature into the broad category of risk assessment and management and sharply distinguish the methods and purposes of that literature from our own. The risk assessment and management literature is concerned with identifying, measuring, reducing, and otherwise managing risk. This includes not only the approaches to risk assessment and perception represented by risk researchers at Chicago, Wharton, Harvard, Colorado, Tufts, and other mainstream research institutions, but also work in the “normal accident” tradition of sociology epitomized by the sociologist Charles Perrow (Perrow 1999).
Across all the disciplines, the risk assessment and management literature has instrumental purposes that the essays in this book do not share. In other aspects of our lives, we applaud many of the efforts and goals of this literature. But our purposes in this book are very different. We are interested here less in what is a risk than we are in the social construction of risk.6 And, in truth, we are less interested in the social construction of risk per se than we are in the use of risk in the social construction of reality. Put perhaps more intelligibly, we are less interested in what is a risk than we are in what is done in the name of risk.
Although the essays in this book differ from one another in many ways, each explores how risk institutions and visions shape the social world. From an opening chapter by Tom Baker analyzing the role of insurance institutions in the distribution of responsibility, through a concluding chapter by François Ewald suggesting that the logic of government may be moving beyond insurance to precaution, each contributor treats the (p.19) current technologies, forms, and visions of risk and insurance as problems in their own right.
Our focus on the use of risk is similar in some ways to work by Mary Douglas and others in what has been described as a cultural approach to risk (Rayner 1992). That work has persuasively argued that what is a risk differs across time and space, not according to an objective, scientific process, but rather according to the logic and influence of institutions: “Whatever objective dangers may exist in the world, social organizations will emphasize those that reinforce the moral, political or religious order that holds the group together” (Rayner 1992:87). In other words, people in organizations actively select among and then use risks to further goals that bear at best a very loose connection to their expressed purpose of assessing and managing the objective danger that exists in the world. For example, in their well-known study of environmentalism, Douglas and Wildavsky (1982) argued that the shift in public concern from national security to environmental risk did not result from a change in the objective risks at issue. Instead, it resulted from a shift in influence from traditional center institutions dominated by markets and bureaucracies (such as large corporations, the military, and government agencies) to sectarian organizations (such as environmental movements and peace groups). The center had placed a higher priority on risks to prosperity and national security, while sectarian organizations emphasized environmental risk.
The contributors to this volume undoubtedly would all ascribe to the core cultural construction of risk insight of this literature. Fundamentally, however, the cultural risk analysts are engaged in a very different project than ours, one that lies much closer to the risk assessment and management project. Particularly in her later work, Douglas and her collaborators are engaged in advising policy makers about how to understand public perceptions of and reactions to the potential dangers that exist in the world (see, e.g., Douglas 1985; Rayner 1992). Our work lies more distant from the “pull of the policy audience” and more focused on social change (Sarat and Silbey 1988).
In addition, we take issue with the exclusive focus of even the cultural risk literature on the avoidance of “risk as danger.” In one of her more recent essays, Mary Douglas states that what is “new” about risk is that it has become a synonym for danger: “Whereas originally a high risk meant a game in which a throw of the die had a strong probability of bringing great gain or great loss, now risk refers only to negative outcomes. The word has been preempted to mean bad risks” (Douglas 1990:3). Risk in this view is (p.20) something to be avoided, spread, or otherwise managed, not something to be encouraged or embraced. Yet this account leaves out the extraordinary amount of risk-seeking behavior in our culture—the stock market, lotteries, sports betting pools, extreme sports, recreational drug use (including tobacco—teens smoke in part because it's dangerous [Ponton 1998])—and the equally significant phenomena of vicarious risk seeking epitomized by what the New York Times called “explornography” (Tierney 1998; Simon, chapter 8 this volume).
We think that risk today is not only about bad risks, but also about opportunity. Many of the phenomena we describe as embracing risk proceed from an implicit belief that risk is a positive force that can be directed toward socially useful ends. Indeed, an awareness that risk can be (and currently is being) socially constructed as a good as well as a bad is among the things that most distinguishes the essays in this volume from much of the earlier risk literature.
Embracing Risk and the Risk Society
Like much of the work collected in this book, scholarship in the “risk society” tradition of Ulrich Beck takes a more encompassing sociological perspective (Beck 1992; Giddens 1990; Ericson and Haggerty 1997). This scholarship chronicles the emergence of a “world risk society” in which exposure to risk becomes a defining characteristic for social groups. Written in a post-Marxist tradition with greater immediate resonance for European than U.S. readers, the risk society scholarship focuses on the emergence of new risks that transcend traditional social boundaries like class and nation. Examples include mad cow disease, acid rain, nuclear power, and the risks that may accompany genetically modified organisms. These differ from many risks of an earlier era because they are global, they threaten the privileged as well as the underprivileged, and they challenge the capacity of traditional insurance institutions.
We share Beck's sense that recent years have witnessed an important change in the way risk is articulated and deployed. Our theme of embracing risk parallels in some ways his argument that concerns about risk are reshaping the dominant strategies of security. Indeed, many of the specific practices described in this volume—new ways of managing mental illness (chapter 9), new forms of sport and recreation (chapter 8), and new uses for insurance technologies (chapter 6)—can be seen as ways of responding to a crisis of governance that Beck's risk society is one account of.
Yet we part ways with Beck in what might seem to be contradictory fashion. On the one hand, we find his concept of risk unduly narrow. It parallels (p.21) the “risk as danger” approach of the risk and culture literature, and misses the appeal of embracing risk.7 On the other hand, we find the scope of his sociological ambition too grand: an example of the “one size fits all” standardization that epitomizes the modernist approach whose decline he chronicles. It is an irony that we are all vulnerable to as well. The Risk Society remains deeply anchored in the sociological tradition of grand theorizing. Yet that tradition is a good example of the modernism that Beck critiques in his work.
We mean our concept of embracing risk to be more provisional. Indeed, as the concluding essay by François Ewald demonstrates, there are competing paradigms at work. The essays in this book explore and challenge the embrace of risk, its historical development, and its contemporary meaning. In the process, they sketch the outline of what we almost hesitate to call (for fear of interrupting a remarkable discussion across the usual lines of academic disciplines and divisions) a sociology of insurance and risk, which would be concerned with an important and underexamined set of institutions and practices that address risk. It would develop a vocabulary to define and describe insurance institutions and their development over time. It would explore their historical and social contexts, their initial purposes, and the process through which they developed purposes and intended and unintended effects of their own. It would also explore how these insurance institutions and practices affected other institutions and practices.
We are not going to set out here the or even a sociology of insurance and risk, because we believe the field requires further mapping before a definitive framework or method can be adequately defended. Rather, this book represents the first collection of essays that can be united under the banner of the sociology of insurance and risk and only the third book in English in the field. (The first two are Zelizer's Morals and Markets  and Heimer's Reactive Risk and Rational Action .) Our aim is to foster, and in some cases, further conversations across disciplines, institutions, and continents. Toward that end, our efforts here are avowedly exploratory and tentative. The best evidence of the promise of a sociology of insurance and risk lies in the chapters written by our twelve contributors. Thus, we encourage you to read on.
Baker, Tom. 1994. “Constructing the Insurance Relationship: Sales Stories, Claims Stories and Insurance Contract Damages.” Texas Law Review 72:1395–1433.
———. 1996. “On the Genealogy of Moral Hazard.” Texas Law Review 75:237–92.
———. 2000. “Insuring Morality.” Economy and Society 29:559–77.
Beck, Ulrich. 1992. Risk Society. London: Sage.
Bernstein, Peter L. 1992. Capital Ideas. New York: Free Press.
———. 1996. Against the Gods. New York: John Wiley.
Burchell, Graham, Colin Gordon, and Peter Miller, eds. 1991. The Foucault Effect: Studies in Governmentality. Chicago: University of Chicago Press.
Burger, Edward J., ed. 1993. Risk. Ann Arbor: University of Michigan Press.
Calabresi, Guido. 1970. The Cost of Accidents. New Haven: Yale University Press.
Dawson, Miles M. 1895. The Principles of Insurance Legislation. New York: Hamboldt Library.
Doran, Nob. 1994. “Risky Business: Codifying Embodied Experience in the Manchester Unity of Oddfellows.” Journal of Historical Sociology 7:131–54.
Douglas, Mary. 1985. Risk Acceptability according to the Social Sciences. New York: Russell Sage.
———. 1990. “Risk as a Forensic Resource.” In Risk, ed. E. J. Burger. Ann Arbor: University of Michigan.
Douglas, Mary, and Aaron Wildavsky. 1982. Risk and Culture. Berkeley: University of California Press.
Ericson, Richard V., and Kevin D. Haggerty. 1997. Policing the Risk Society. New York: Oxford University Press.
Ewald, François. 1986. L̓Etat Providence. Paris: Grasset.
———. 1991. “Insurance and Risk.” In The Foucalt Effect: Studies in Governmentality, ed. G. Burchell, C. Gordon, and P. Miller. Chicago: University of Chicago Press.
Foucault, Michel. 1977. Discipline and Punish. New York: Pantheon Books.
———. 1978. The History of Sexuality. New York: Pantheon Books.
Giddens, Anthony. 1990. The Consequences of Modernity. Stanford: Stanford University Press.
———. 1999. “Risk and Responsibility.” Modern Law Review 62:1–10.
(p.24) Gordon, Colin. 1991. “Governmental Rationality: An Introduction.” In The Foucalt Effect: Studies in Governmentality, ed. G. Burchell, C. Gordon, and P. Miller. Chicago: University of Chicago Press.
Graham, John D., and Jonathan Baert Weiner, eds. 1995. Risk vs Risk. Cambridge: Harvard University Press.
Hacking, Ian. 1990. The Taming of Chance. Cambridge: Cambridge University Press.
Heimer, Carol A. 1985. Reactive Risk and Rational Action: Managing Moral Hazard in Insurance Contracts. Berkeley: University of California Press.
Kimball, Spencer. 1992. Cases and Materials on Insurance Law. Boston: Little, Brown.
Krakauer, John. 1997. Into Thin Air. New York: Vintage.
Krimsky, Sheldon, and Dominic Golding. 1992. Social Theories of Risk. Westport, Conn.: Praeger.
Lipsky, Michael. 1980. Street Level Bureaucracy: Dilemmas of the Individual in Public Services. New York: Russell Sage Foundation.
LoPucki, Lynn M. 1996. “The Death of Liability.” Yale Law Journal 106:1–92
Lowi, Theodore J. 1990. “Risk and Rights in the History of American Governments.” In Risk, ed. E. J. Burger. Ann Arbor: University of Michigan Press.
Margolis, Howard. 1996. Dealing with Risk. Chicago: University of Chicago Press.
Nonet, Philippe. 1969. Administrative Justice: Advocacy and Change in a Government Agency. New York: Russell Sage Foundation.
O̓Malley, Pat. 1991. “Legal Networks and Domestic Security.” Studies in Law, Policy and Society 11:171–90.
———. 2000. “Uncertain Subjects: Risks, Liberalism and Contract.” Economy and Society (forthcoming).
Pal, Leslie A. 1986. “Relative Autonomy Revisited: The Origins of Canadian Unemployment Insurance.” Canadian Journal of Political Science 19:71–92.
Perrow, Charles. 1999. Normal Accidents: Living with High-Risk Technologies. New York: Basic Books.
Piven, Francis Fox, and Richard Cloward. 1971. Regulating the Poor: The Functions of Public Welfare. New York: Pantheon Books.
Ponton, Lynn E. 1998. The Romance of Risk: Why Teenagers Do the Things They Do. [[Au: city?]]: Perseus Books.
Rayner, Steve. 1992. “Cultural Theory and Risk Analysis.” In Social Theories of Risk, ed. S. Krimsky and D. Golding. Westport, Conn.: Praeger.
Reichman, Nancy. 1986. “Managing Crime Risks: Toward an Insurance-Based Model of Social Control.” Research in Law, Deviance and Social Control 8:151–72.
Rees, Joseph. 1994. Hostages of Each Other: The Transformation of Nuclear Safety since Three-Mile Island. Chicago: University of Chicago Press.
Rose, Nikolas. 1999. Powers of Freedom. Cambridge: Cambridge Univeristy Press.
Russo, Richard. 1988. The Risk Pool. New York: Random House.
Sarat, Austin, and Susan Silbey. 1988. “The Pull of the Policy Audience.” Law and Policy 10:97–166.
(p.25) Simon, Jonathan. 1988. “The Ideological Effects of Actuarial Practices.” Law and Society Review 22:771–800.
———. 1993. Poor Discipline: Parole and the Social Control of the Underclass, 1890–1990. Chicago: University of Chicago Press.
———. 1994. “In the Place of the Parent: Risk Management and the Government of Campus Life.” Social and Legal Studies 3:14–45.
———. 1998. “Driving Governmentality: Automobile Accidents, Insurance and the Challenge to Social Order in the Inter-War Years, 1919–1941.” Connecticut Insurance Law Journal 4:521–88.
———. 1999. “Law after Society.” Law and Social Inquiry 24:143–94.
Skocpol, Theda. 1992. Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States. Cambridge: Belknap Press, Harvard University Press.
Smelser, Neil J., and Richard Swedborg, eds. 1994. The Handbook of Economic Sociology. Princeton: Princeton University Press.
Squires, Gregory, ed. 1997. Insurance Redlining. Washington: Urban Institute Press.
Starr, Paul. 1982. The Social Transformation of American Medicine. New York: Basic Books.
Stenson, Kevin, and Paul Watt. 1999. “Governmentality and ‘the Death of the Social’?: A Discourse Analysis of Local Government Texts in South-east England.” Urban Studies 36:189–201.
Stone, Deborah. 1994. “Promises and Public Trust: Rethinking Insurance Law through Stories.” Texas Law Review 72:1435–46.
Tierney, John. 1998. “Explornography: The Vicarious Thrill of Exploring When There's Nothing Left to Explore.” New York Times Magazine, 26 July, 18.
U.S. Department of Health and Human Services. 1994. “Availability of Funds for School Health Services and Health Education/Promotion for Homeless and At Risk Children and Youth.” Federal Register 59:24171.
Zelizer, Viviana A. 1979. Morals and Markets: The Development of Life Insurance in the United States. New York: Columbia University Press.
———. 1985. Pricing the Priceless Child: The Changing Social Value of Children. Princeton: Princeton University Press. (p.26)
(1.) Our primary focus in this introduction is on the developments in the United States, but the trends we describe, both past and present, apply with some differences to other postindustrial societies. A number of the essays in this book examine other countries, including France, the United Kingdom, and Canada. While the differences (p.22) among these societies are worthy of separate treatment, for now we want to stress the commonalities.
(2.) Ewald's assertion that contemporary culture is becoming more “riskphob[ic]” (Ewald, chapter 11 this volume) might seem to contradict the embracing risk thesis. Yet, our views in fact are largely consistent, differing perhaps most in the national mood about these changes in our two respective countries (France and the United States). We agree with Ewald that the risk-spreading (or solidarity) paradigm is on the wane, though we differ in the emphasis we place on the various responses to that trend. Ewald mainly focuses on catastrophic risks, for which a new mandate of precaution may be replacing the previous emphasis on prevention and compensation. Much of what we have described as embracing risk operates in the arena of noncatastrophic risk. Indeed, many of the private insurance market adaptations to the embrace of risk encourage individuals and institutions to maintain catastrophic insurance coverage. In an important sense our “embracing risk” and Ewald's “precaution” operate at two different ends of the probability spectrum. Embracing risk makes most sense for risks that are relatively predictable and capable of being managed by individuals and organizations. On the other hand, precaution makes most sense for risks that are unpredictable (indeed, in that sense would not be regarded as “risks” at all within the traditional, and not always sustainable, distinction between risk and uncertainty). At the same time Ewalds's essay may also reflect a genuine difference in atmosphere between France and the United States. In Europe, with a stronger environmental movement and less hostility toward state regulation, there does seem to be a more pronounced shift toward precaution than in the United States. (Though the public aversion to low-probability, high-consequence risks explored by Margolis (1986), for example, may be the U.S. version of the move to precaution.)
(3.) This is one of the places where we have varied Ewald's framework. He appears to have a more fixed notion of insurance technology than we wish to adopt. In his 1991 essay, he identifies “the” technology of insurance as the art of combining risks, using the term risk in a somewhat technical sense of future harms that can be addressed through the language of probability. We regard insurance technology as both a larger and a smaller category than that. Insurance technology is a larger category in the sense that it can address nonprobabilistic risks (which would not be “risks” at all within Ewald's definition of risk). Insurance technology is a smaller category in the sense that we are at least as interested in the individual components involved in combining risks, which differ among insurance institutions and forms and over time and place, as we are in what Ewald calls the “art of combinations,” which most insurance institutions share.
(4.) This is another place where we have varied Ewald's framework. The English translation of his work uses the word imaginary to refer to this category. We decided against adopting that term because of its well-entrenched, somewhat negative connotation of “counterfactual.” Although the word has some appeal because it emphasizes the role of imagination in culture, we substituted a term that we believe will be less likely to confound understanding: “vision.” Regardless whether “imaginary” or “vision” is the better term, your thought process at this moment justifies our disagreement with Ewald, because you have stopped to consider the role of imagination in constructing social meaning.
(5.) We must be clear from the very start that this in itself is not a criticism. Following Foucault (1977) we take for granted that all ways of gathering knowledge require access to the channels and flows that exercises of power create, and that all ways of exercising power require methods for producing knowledge about the subjects of power.
(6.) For that reason, we deliberately have almost nothing to say about the low-probability, (p.23) probability, high-consequence risks that have received so much attention in the risk studies literature (see, e.g., Margolis 1996) and, thus, leave aside the interesting and important psychometric and social cognition issues raised by those risks.
(7.) One difference here, paralleling some of our differences with Ewald, may be cultural. It may be in this regard that national differences in interpreting these developments are particularly pronounced. France, Germany, and Japan, among other nations, remain more politically committed to risk spreading while the United States, the United Kingdom, Australia, and New Zealand have been most enthusiastic in embracing risk. For example, concerns about high technology and biomedical risks have been far more prominent on the European political agenda then in the United States.