In Averting Catastrophe, Cass Sunstein continues his lengthy body of work on decision theory and risk management in public policy. He focuses in this book on regulatory decisions under “Knightian uncertainty,” where the probabilities of different outcomes are truly unknown. Under these conditions of uncertainty, Sunstein endorses use of the “maximin” principle, which recommends choosing the option with the “best” minimum outcome, thus avoiding the greatest risk. Specifically, he suggests use of the maximin principle under the following limited conditions: (1) near-total uncertainty about the probability of different outcomes, (2) relatively low costs for preventing the worst-case outcome, and (3) relatively high social costs of the worst-case outcome. By focusing on regulatory choices under such conditions of deep uncertainty, the book makes a helpful contribution to the burgeoning debate over regulatory policy regarding environmental, public health, and other complex risks. At the same time, the book is less effective in applying some of its abstract ideas to concrete examples such as climate change or COVID-19, despite its subtitle. Nevertheless, this is a carefully argued and concise treatment of important issues in risk management policy, one that would be useful for those studying or teaching risk management policy and decision making under conditions of uncertainty.

The book starts by discussing the difference between conditions of “risk,” where we know the probabilities of different outcomes of our decision, and conditions of “uncertainty,” where we do not know those probabilities. Sunstein makes a convincing argument about the existence of situations where probabilities of different outcomes are truly unknown and their applicability to some important if “unusual” policy dilemmas (p. 80). This defense of uncertainty as an important condition for regulatory decisions is a welcome foray into an area that would benefit from more attention as governments face ever more complex and unpredictable threats.

Sunstein outlines the basic idea of the maximin principle through a set of hypotheticals, including cases where the rule seems consistent with our intuition and others where it does not. For example, Sunstein describes how the maximin rule conflicts with most people’s intuition by recommending we choose a “50% chance of gaining $50 and a 50% chance of losing $50” over “a 99.9% chance of gaining $10,000 and a 0.1% chance of losing $60” (p. 28). Based on these kinds of stylized dilemmas, Sunstein makes his basic case for a regulatory policy that tries to maximize the “expected value” of possible outcomes, weighting their expected net benefits by the probability of each outcome. This is the heart of his long-standing argument in favor of using cost-benefit analysis and similar tools to guide regulatory standards, an argument he has advanced in his books (e.g., Risk and Reason, 2002; Laws of Fear: Beyond the Precautionary Principle, 2005; The Cost-Benefit Revolution, 2018; and with Richard H. Thaler, Nudge: Improving Decisions about Health, Wealth, and Happiness, 2008) and with some political controversy in a stint as administrator of the Office of Information and Regulatory Affairs.

Sunstein is more sympathetic to using maximin when probabilities for different possible outcomes are unknown, such as a choice between “a 100% chance of losing $1 or…a 100% chance of gaining $100,000, alongside a chance of dying.” Here he argues (p. 67) that maximin’s prescription to avoid the uncertain chance of dying is what rational people might choose.

But even facing uncertainty, Sunstein is careful to press for obtaining more information—a “crude version of cost-benefit balancing” as he puts it (p. 110)—regarding exactly “how bad” the worst outcome is relative to others and for better understanding the opportunity costs of choosing the path that avoids the worst outcome. This is a welcome approach: it defends some commonsense principles about attending to opportunity costs and the scale of the worst-case outcomes where that information can be determined, even when we cannot optimize our decision in the manner of traditional benefit-cost analysis. Coming from a cost-benefit analysis advocate such as Sunstein, this message is even more helpful in moving forward the debate over good regulatory policy.

At the same time, the book is less satisfying in its application of these ideas to the concrete cases likely to attract many readers: climate change and COVID-19. It spends a good deal of time discussing climate change regulations as a case of decision making under uncertainty, but it does so in a manner that feels somewhat out of date. Many experts would question some of the book’s assertions about climate change, such as this quote from MIT economist Robert Pindyck that “we do not even know the probability distributions for future temperatures and impacts” related to climate change (p. 8). If this statement was accurate in 2010 (when this quote from Pindyck was published), it seems dubious today when the world’s leading climate scientists issue increasingly certain statements about the “catastrophic” results of continued global warming trends (IPCC, “Summary for Policymakers” in Climate Change 2013: The Physical Science Basis, edited by T. F. Stocker et al., 2013). Today, there is no uncertainty about the disastrous effects of future global mean temperature increases of 5 degrees Celsius or higher, as expected under current “business as usual” scenarios. Thus, the book’s identification of climate change as a problem of deep uncertainty seems out of date, given the increasingly certain scientific warnings of greater storms and droughts, ocean acidification, sea-level rise, and temperatures too high to allow for safe strenuous outdoor activity if human emissions responsible for climate change continue on their current trends (e.g., IPCC, “Summary for Policymakers” in Climate Change 2013; USGCRP, Impacts, Risks, and Adaptation in the United States: Fourth National Climate Assessment, Volume II, edited by D. R. Reidmiller et al., 2018).

The book’s discussion of the response to COVID-19 in the early stages of the pandemic as an example of policy making under uncertainty is more convincing. Indeed, a more thorough discussion of government responses to COVID-19 using Sunstein’s framework would have been a welcome addition to the text. Although the book clearly identifies the COVID outbreak as a problem where following maximin principles was the right choice (p. 116), it does not delve into the details of what following maximin means for controversial COVID responses. For example, does the rule justify even more aggressive stay-at-home orders than were imposed in some nations, despite the high costs of such actions? Or simply strict masking and social distancing policies? Sunstein does not grapple with these details, which is disappointing for readers hoping for more insight on how his principles would or should guide policy making on this critical issue.

There are other ideas in this book that deserve more attention as well. Sunstein briefly raises the issue of how to deal with “moonshots,” or low-probability chances of “producing extraordinary returns” (p. 37). This is another thought-provoking insight grounded in Sunstein’s long-standing interest in the importance, and challenge, of attending to risks both from overregulation and underregulation, and it deserves more attention that it gets in this slim volume. At the same time, Averting Catastrophe grapples with multiple big issues in an admirably concise text, so perhaps it is unavoidable that it covers some of those issues insufficiently. One can only hope that the author considers these other concerns and issues in future volumes and papers, even as one hopes that regulators will also pay attention to the general principles articulated in the book related to governing under uncertainty.