Leo Tolstoy opened Anna Karenina by reminding us that families can be unhappy in many different ways. Adrian Vermeule devotes his new book, The Constitution of Risk, to reminding us that constitutions can go wrong in many different ways, and that obsessing over just one of these ways is a serious mistake. He does this very effectively, even if the lessons the book imparts are primarily negative. In other words, it tells us quite a bit about how not to design and interpret constitutions, and less about how to perform these tasks correctly.

Vermeule seeks to defend two claims. The first is that “constitutional rulemaking is best understood as a means to regulate and manage political risks” (p. 1). Ordinary government policymaking deals with first-order risks, which are many and varied in nature; constitutional law concerns second-order risks, the risks that arise from the creation of a political process to deal with first-order risks (pp. 3–4). To design a constitution is thus to adopt a certain strategy for dealing with political risks. But what strategy should constitution designers adopt? Answering this question generates the second major claim he defends in his book.

Vermeule contrasts two approaches to risk management, and defends one of them—optimizing constitutionalism—against the other—precautionary constitutionalism (p. 10). (Note that this contrast between approaches only makes sense if Vermeule’s first claim—that constitutions are fundamentally about managing political risk—is true.) “Optimizing constitutionalism,” he writes, “trades off all relevant political risks, giving them their due weight in the circumstances, without any systematic skew or bias against any particular type of political risk” (p. 10). Precautionary constitutionalism, by contrast, holds “that constitutional rules should above all else entrench precautions against the risks that official action will result in dictatorship or tyranny, corruption and official self-dealing, violations of the rights of minorities, or other political harms of equivalent severity” (p. 11). (Vermeule admits, of course, that these two approaches “lie on a continuum”; see pp. 15, 87.)

In making the case for optimizing constitutionalism, the author forcefully argues that prioritizing one political value above all others can generate serious political pathologies. This argument is very persuasive and is the strongest part of the book. The most serious political pathology he identifies is the “pervasive and frequently irrational attitude of distrust of official power that is so prominent a feature of American constitutional history and discourse” (p. 24). In response, he endorses Madison’s point in Federalist 63 that “Liberty may be endangered by the abuses of liberty as well as by the abuses of power; that there are numerous instances of the former as well as of the latter; and that the former, rather than the latter, are apparently most to be apprehended by the United States” (emphasis in original; quoted at p. 68). In the United States today, the Tea Party appears to exist for no other reason than to prove Madison’s point.

While Vermeule’s case against prioritizing one political value (and the avoidance of one type of political risk) above all others is compelling, I am not convinced that the distinction between optimizing and precautionary constitutionalism is the right way to make this point. I say this for three reasons. First, many of his examples do not work quite as intended. As noted before, Vermeule quotes Madison as recognizing the trade-off between abuses of liberty and abuses of power; it was recognition of this tradeoff that led Madison to endorse a powerful but limited federal government. Vermeule also notes that during Virginia’s debates over constitutional ratification, Madison “objected to the supposition that ‘the general legislature will do everything mischievous they possibly can’” (p. 31, quoting Jon Elster). And yet strangely, Vermeule deems Madison a precautionary constitutionalist. Perhaps this is because he takes Alexander Hamilton—a man who never met a power he did not want to assign to the national government—to be an exemplar of optimizing constitutionalism (p. 20). I would regard Hamilton as a precautionary constitutionalist obsessed with preventing excesses of government weakness, not government strength; this is the opposite of the usual American obsession, to be sure, but it seems just as far away from the “mature position” promoted by Vermeule (p. 13).

Second, in making his case against precautionary constitutionalism, Vermeule conflates two different moves one could make in constitutional design. On the one hand, he associates precautionary constitutionalism with the view that “constitutional rulemakers and citizens design and manage political institutions with a view to warding off the worst case” (p. 11). On the other hand, he declares that “precautionary constitutionalism focuses myopically on just a particular target risk or set of target risks, while neglecting countervailing risks” (p. 14). In contrast, “[o]ptimizing constitutionalism holds that all relevant political risks matter, and that a systematically precautionary and distrustful approach to the constitutional allocation of power is a mistake” (p. 13). But it is one thing to say that constitutional designers should take all relevant political risks into account; it is quite another thing to say what attitude constitutional designers should adopt toward those risks. Surely one could, without contradiction, argue that constitutions should be designed with the best worst-case scenario in mind, where the worst case is defined by reference to all possible sources of risk. Vermeule has no patience for “maximin constitutionalism,” which he regards as an “extreme form of the precautionary approach” (emphasis in original; p. 11). But he never considers the possibility of dissociating the maximin decision-making rule from the myopic focus upon single sources of political risk that he (quite reasonably) opposes.

Third, while Vermeule’s injunction to heed all forms of political risk is very sensible, this injunction by itself does not yield a distinct approach to constitutional design and interpretation. Optimizing constitutionalism, as he defines it, does not seem to provide such an approach. Optimizing constitutionalism “does not necessarily entail cost-benefit analysis,” unless perhaps such analysis is understood as “a loose synonym for consequentialism” (pp. 78, 79). Indeed, by its very nature, optimizing constitutionalism does not fit well with the idea of maximizing well-being, or economic value, or anything else, for that matter; if constitutional designers focused upon maximizing one political value, then they would be singling out some forms of political risk and ignoring others—precisely the mistake Vermeule wishes everyone to avoid. Moreover, maximization requires the assignment of probabilities to outcomes; the author, in contrast, believes “that genuine uncertainty does sometimes exist, and that the strict Bayesian approach is misguided” (p. 7).

As a result, optimizing constitutionalism tells you much about how not to design constitutions and very little about how to design them. Vermeule admits as much: “The only general positive counsel I offer, then, is that constitutional rulemakers should consider all relevant risks, on all sides of relevant questions. As a positive counsel this is obviously thin, even banal. But its value lies in its negative aspect. It is a placeholder for, and a summary reminder about, a rich variety of ways in which constitutional rulemakers can go badly wrong. … At bottom, optimizing constitutionalism offers a master principle of avoiding obsession with particular risks, and it arms rulemakers with a breviary that lists and illustrates the ways in which such an obsession may come into being” (emphasis in original; pp. 14–15).

I do not accept all of the illustrations Vermeule offers for his argument—it would be surprising if I did, given how many illustrations he offers (it is a very rich book)—but I completely agree that the master principle he recommends belongs in any constitutional designer’s toolbox. I cannot help observing, however, that above all else, Vermeule’s master principle is—precautionary in nature.