This is the era in which many prefer to be safe than to be sorry. In the past five years, we have seen the United States justify a preemptive war against Iraq lest Saddam Hussein use weapons of mass destruction against us or others. Ordinary Asian citizens donned surgical masks to avoid contracting severe acute respiratory syndrome, or SARS. And in the wake of the terrorist attacks on September 11, 2001, Americans stocked up on duct tape so that they could be “sheltered-in-place.” Is not an ounce of prevention worth a pound of cure?

In this tour de force, Cass R. Sunstein persuasively puts the lie to the increasingly popular concept known as the “precautionary principle” (PP). According to the PP, “regulators should take steps to protect against potential harms, even if causal chains are unclear and even if we do not know that those harms will come to fruition” (p. 4). As President George W. Bush put it when justifying the preemptive war in Iraq, “ ‘If we wait for threats to fully materialize, we will have waited too long,’ ” and “ ‘it is essential that when we see a threat, we deal with those threats before they become imminent. It's too late if they become imminent’ ” (p. 4). The problem with this mode of thinking, argues Sunstein, is that it is literally incoherent. Regulation may well protect against potential harms, but failing to regulate may protect against potential harms, too. After all, it was risky both to allow Iraq to threaten use of its weapons and for the United States to start a war we did not know how to conclude. If we were true to the PP and aimed to avoid all risks, we could neither go to war nor refrain from war.

In Chapters 1 and 2, Sunstein vitiates the PP as a matter of both theory and practice. He then goes on in Chapters 3 and 4 to explain why—and how—it nevertheless appeals to so many people. Part of the explanation stems from our tendency to focus on the worst-case scenario, even if it is unlikely. He describes this as the phenomenon of “probability neglect” (pp. 39ff), which becomes more salient when people feel strongly about the possible bad outcome. Part of the explanation comes from the social context in which fear operates. As Sunstein puts it, fear is contagious. It can spread through a series of “social cascades” (pp. 94ff), which can extend the fear geometrically; simply put, each person shares his or her fear with several others, who in turn infect several others, who frighten still others, and so on. Intense fears of bad outcomes can also intensify when groups of people confer about their worries. When people get together to discuss their fears, rumors spread, fears grow more extreme, and people end up more afraid than they were at the outset. The author calls this phenomenon “group polarization” (pp. 98ff) and uses it to explain how moral panics happen. In both chapters, he deftly substantiates his claims with a careful analysis of evidence from psychology, cognitive science, policy studies, current events, and history.

In the remainder of the book, Sunstein provides “solutions” to the problem of irrational fears. In Chapter 5, he introduces the “Anti-Catastrophe Principle” for managing risks to which we cannot assign a realistic probability. This principle is much narrower than the PP, calling on us to minimize the worst-case scenario. We should only do this, he argues, after first evaluating fully the associated social risks, striving to be cost-effective, ensuring that the risks are distributed fairly, and paying close attention to the costs (and people's willingness to pay). People may panic when contemplating “irreversible harms” (pp. 115ff), but we need not act on those fears unless the predicted harm is sufficiently probable and of appropriate magnitude.

Sunstein recognizes that government now manages risk through the use of cost–benefit analysis (CBA). After providing an outstanding summary of how CBA actually works in Chapter 6 (including a fascinating description of how the federal government calculates the “uniform value for a statistical life” [pp. 132ff]), he explains why this flawed method is nonetheless necessary. In short, CBA can help “discipline” the analysis of risks and promote “coherence,” even as it translates into dollar figures the otherwise nonmonetary values associated with (not) regulating risks (p. 174). He wisely modifies CBA so that it allows for consideration of people's willingness to pay to alleviate the feared risks. He also advocates choosing regulatory options that may not maximize net benefits if those policies provide the best way to avoid harming the disadvantaged or violating rights.

On my reading, these are welcome alterations to a necessary but deeply flawed policy tool. As Elizabeth Anderson has so eloquently argued in Value in Ethics and Economics (1993), CBA evaluates all costs and benefits monetarily, but we often value things using alternative modes of valuation that do not translate easily into dollar figures. This means that CBA either excludes important values from consideration or includes them in a way that fails to capture their importance to us. For example, how can we put a “price” on the value of life, rights, or equitable distribution of costs and benefits? Moreover, even Sunstein's modified CBA would be done by bureaucrats who are not electorally accountable. This may well allow for rational management of the public's irrational fears, but removing decision making from the democratic process is, prima facie, problematic. As a leading deliberative democrat, Sunstein acknowledges this problem and advocates public education on real and exaggerated risks. This certainly helps to create conditions where deliberation could be more rational. But he is also right that when fear escalates into a contagious irrational fear, it undermines deliberative democracy. He makes a strong case for tolerating a potentially undemocratic solution to facilitate more genuinely deliberative democracy, but it is a bitter pill to swallow, especially for those who would rather be safe than sorry.