Nudge the judge

Cass Sunstein believes judges can nudge.

Using statistical analogies, “behavioral law and economics” claims to improve on neoclassical law and economics by having a higher R squared, and to make predictions with fewer unrealistic restrictions (i.e. fewer degrees of freedom). (Jolls et al., 1998, at 1487, 1489) [i]

The widespread existence of laws that are neither efficient nor an obvious result of rent-seeking by powerful groups is difficult to explain within the neoclassical framework. Behavioralists supplement this account with reference to bounded rationality and bounded self-interest (in the form of fairness norms). Inefficient bans on ticket scalping, for instance, exist because of fairness norms, and most people judge fairness by how closely it corresponds to a non-scalped price.

While undoubtedly filling in key details of economic agents’ internal environment, behavioralists end up having to create a whole host of exogenous events. From Austrian thinking, behavioralists emphasize that fairness norms must be examined before the law came into being, since the existence of law itself shapes fairness norms. As in Austrian thought, however, this is difficult to falsify or investigate. Following neoclassical thought, Sunstein and colleagues assume that legislators are maximizers interested in their own reelection, and so respond to voter fairness concerns. They also assume the presence of “availability entrepreneurs” that manipulate cognitive biases by reminding voters or legislators of catastrophes they wish to regulate. (What motivates these entrepreneurs is not explored.) So the mechanisms causing shifts in cognition and translating those shifts into policy are left outside of the theory. (Jolls et al., 1998, at 1508-1519)

Given the role played to exogenous forces, it is curious that Sunstein and colleagues choose to examine Superfund toxic waste dump legislation as a case study. The authors maintain that the legislation is not justifiable on cost-benefit grounds, and outside interest groups weren’t even advocating for it. Instead, it was solely because availability entrepreneurs within the Environmental Protection Agency attempted to make the issue more salient with the public. This example raises more questions than it answers. What were these government bureaucrats attempting to maximize? The authors suggest that they were “concerned about an apparent gap in federal law and eager to consolidate their authority over issues of public health.” (Jolls et al., 1998, at 1520-22) But how do state agents develop autonomous interpretations of social reality? What are they influenced by or attempting to maximize? To answer these questions, you would need a theory of the state. (Evans, 2011) (Chang, 2002)

Likewise, the behavioral account leaves unclear what exactly judges do, and what they maximize.

Like in Posner, the judge or “nudger” in behavioral economics pushes disputants towards superior outcomes. In the former, this is a move towards efficiency by simulating market conditions in high transaction cost settings. In the latter, it is a move towards procedural rationality by helping disputants overcome biases. In Posner, judges play this role because they are selfish optimizers who want to maximize citations. In behavioralism, we don’t know what compels judges to take on this function.

Elsewhere, Sunstein has argued that judges also have bounded rationality, and therefore satisfice through partial, abstract and very open-ended rulings. Constitutional constraints and their professional inability – unlike administrative law staff – to engineer social changes also play a role in limiting the domain of judges in social life. (Rubin, 1997, at 282-285) But this approach fails to explain the emergence of judicial institutions and norms in the first place.[ii]

(Indeed, the radical indeterminacy of behavioral law and economics is best captured by this sentence: “bounded rationality may increase the need for law (if government’s failings are less serious than citizens’), [but] bounded self-interest may reduce it, by creating norms that solve collective action problems even without government intervention.” (Jolls et al., 1998, at 1541))

Unable to provide (much of) a positive account, Sunstein and company turn to prescription and normative analysis. Because of the cognitive biases (i.e. hindsight bias, revenge, etc.) affecting human cognition, societies should rely less on juries and more on administrative agencies. When this is not possible, judges and lawyers should withhold information from jurors in order to neutralize their cognitive defects.[iii] While noting the possible ethical objections to intentionally distorting reality (not to speak of the problems for accurate and utility maximizing private ordering), they argue that (as in the case of bounded willpower) individuals often limit their own choice and information set, so “nudgers” should be able to as well. (Jolls et al., 1998, at 1522-1532) (In subsequent work, the authors focus primarily on how behavioralism could be harnessed for legal engineering ends, such as through instructions to juries to not be “Monday morning quarterbacks” [thereby reducing hindsight bias] or requiring litigants to make the other side’s arguments in court [thereby reducing the self-serving bias] (Jolls and Sunstein, 2006, at 236)) In short, the neoclassical account finds that law corrects market failure, while the behavioral account find that it corrects cognitive ones.

Other behavioralists have looked more specifically at judging. (Cohen and Knetsch, 1992) argue that judges tend to reflect the average member’s intuition of what’s fair. So the loss aversion that behavioralists observe is reflected in the case law.

(Rachlinski, 1998) offers a similar account. He notes that “Courts’ ex post judgments of ex ante decisions fall into three categories: (1) judgments under objective (‘should have known’) standards; (2) judgments under subjective (‘did know’) standards; and (3) judgments of what was foreseeable…. When a court must determine what someone ‘knew or should have known,’ it is especially likely to fall prey to the hindsight bias. Consider the assessment of negligence in a tort action as an example. A defendant’s conduct was negligent if it created an ‘unreasonable risk of harm’ to others, ‘which the actor should recognize at the time of his action or inaction.’” The hindsight bias plays a significant role in the ex post assigning of ex ante probabilities to outcomes. Rachlinski suggests that hindsight bias plays less of a role in the subjective and foreseeability judgments, since the former is about documenting what was known, and the latter is not about assigning probabilities but is a qualitative assessment about possible (not probable) future states of the world that can be quite tough on defendants with or without the benefit of hindsight. (Rachlinski, 1998, at 590-591)

Nonetheless, the author suggests that the law already responds efficiently to the bounded rationality problem. As an example, he cites the “business judgment rule,” whereby judges (in order to overcome their own hindsight biases), will not hold a corporate executive responsible for negligence if the latter “rationally believes that his business judgment is in the best interests of the corporation.” The justification for this relatively lax standard of review is that courts are said to not know as much as corporate boards about how to best run a company. The same deference is not shown to lawyers and accountants, but law and economics explains this by noting the economic justification of not wanting capitalists to take an “excess of precautions”, and because shareholders only lose their future dividends while lawyer or accountant negligence may cost a non-equity holder some sunk amount. As Rachlinski writes,

“Without the hindsight bias, it is difficult to explain why corporate managers are immune from liability for negligence and other professionals are not. Arguably, all professionals have sufficient incentives to avoid negligent conduct-they need to do so in order to retain their patients or clients. Therefore, all professionals could argue that they should be immune from liability for negligent conduct. Alternatively, professionals have much more information than their clients, patients, and shareholders, and are in the best position to reduce the social costs of their actions. Therefore, all professionals (corporate managers included) should be subject to liability. When negligence judged in hindsight is revealed to be a system of quasi-strict liability, however, the distinction becomes clear. Strict liability is a perfectly defensible, even laudable, form of liability-it creates incentives to be efficient and spreads the risk of injury, reducing the possibility that a single individual will experience a catastrophic loss. But adopting a system of strict liability raises the costs of any activity, which might create some unwanted consequences. The consequences of a system of strict liability are particularly undesirable to shareholders in a corporation. Shareholders already spread their risks through diversification and, even without a system of quasi-strict liability, they must worry that their managers make risk-averse decisions. With decades of experience in the matter and a basic understanding of the foibles of judging in hindsight, courts seem to have realized this, and carved out the business judgment rule for corporate managers alone among professionals.” (Rachlinski, 1998, at 619-623)[iv]

As this quote makes clear, neoclassical presumptions of legal efficiency can be paired with behavioral insights to argue for the superiority of an otherwise arbitrary and politically determined system that insulates the very rich relative to middle class professionals.

Finally, in a non-randomized sample of judges that had attended a law and economics standard, (Viscusi, 1999) found that judges were less likely to have hindsight bias than a panel of mock jurors. He suggests this may be because of their higher level of education, and ability to put risky decisions into context.