Clarity Doctrines

Introduction

Legal practice is riddled with claims about when the law is or isn’t “clear.”1 If a statute is unclear or ambiguous, a court might defer to an agency,2 side in favor of lenity,3 or avoid interpretations that would render the statute unconstitutional.4 And when a question of constitutional law is clear or open to only one reasonable view, a court might conclude that a government officer can be held liable for damages5 or that a state court judgment of conviction must be overturned.6 So legal clarity is widespread; and when it doesn’t exist, we find unclarity, ambiguity,7 or reasonable disagreement.8

Yet there is a growing realization that the legal system is unclear about the meaning of legal clarity. And critics, including Justice Brett Kavanaugh, have forcefully argued for eliminating or curbing reliance on clarity.9 Courts often use similar or interchangeable terminology when they discuss clarity under Chevron, federal habeas corpus, qualified immunity, lenity, and avoidance, among other doctrines. But courts have paradoxically left the nature of legal clarity underspecified.10 How are judges supposed to know when a legal view is clear enough to trigger any of the aforementioned doctrines?11 Because case law affords no definitive answer to that basic question, courts cannot begin to address more sophisticated issues, such as whether clarity thresholds are (or should be) constant across different cases, to say nothing about different doctrines.12 Ambiguity’s evident ambiguity also raises rule-of-law concerns: If courts cannot articulate just what qualifies as legal clarity, can they really know it when they see it?13

This Article attempts to clarify the nature and proper design of “clarity doctrines,” that is, legal principles that incorporate the idea of legal clarity or its absence. That effort begins with a simple but fundamental point: a judicial finding that a legal issue is clear or unclear is not just a declaration about a preexisting state of the world, but rather a consequential action within a legal system. Who wins and who loses often depends on what a court thinks about the law’s clarity. Judicial declarations that a statute is ambiguous, that a legal view is reasonable, and so forth are thus actions that require justification. By contrast, courts sometimes assume that legal clarity is reducible to a linguistic state and therefore purely empirical—in which case it would either exist or not, no matter what the law called for, much as uranium exists on the earth no matter what the Constitution or any judge might say. But once we see that findings of legal clarity are judicial actions, our focus naturally shifts to the possible justifications for those actions, including justifications grounded in law. True, the law will sometimes dictate that a clarity finding should turn on an empirical question, such as the communicative content of a particular statutory text.14 But even then, the choice to care about clarity or unclarity, as well as how to go about finding it, would still be governable by law. In short, the law can and should establish criteria for legal clarity as well as the related concepts of ambiguity and reasonable disagreement.

But what should the law of clarity be? The answer depends on the doctrine at issue. In general, the law should direct findings of legal clarity when doing so advances legally recognized goals. Moreover, doctrinal goals can, and often do, vary by context.15 Consider the important distinction commentators have drawn between “internal” and “external” assessments of ambiguity.16 Internal assessments focus on an interpreter’s confidence about her own views of the right answer to a legal question, whereas external assessments refer to an interpreter’s confidence about what other interpreters would think is correct. Different clarity doctrines have good reason to care primarily about one or the other way of assessing ambiguity. For instance, if the rule of lenity aims to give notice to potential offenders, then its applicability in any given case ought to turn on how certain other interpreters—namely, potential offenders—would answer legal questions.17 That legal objective thus directs judicial attention toward a particular type of externally assessed ambiguity. By contrast, doctrines that are less concerned with notice, and more concerned with getting the right answer, might call for courts to give greater attention to internal assessments of ambiguity—that is, their own best view of the matter.

The distinction between internal and external perspectives points toward two model forms of legal clarity, and the choice between them turns on the goals of the doctrine at issue. First, clarity as certainty adopts the perspective of the deciding court and asks whether the court is confident of the right answer. The court may consider accuracy-promoting heuristics and account for the views of comparably expert or like-minded interpreters. That approach makes sense when accuracy is at a premium. But because judges tend to view their beliefs as not just correct but clearly correct, clarity as certainty runs the risk of collapsing into an investigation of the merits. Chevron deference offers an example. Second, clarity as predictability adopts the perspective of an actor apart from the deciding court and makes allowances for that actor’s distinctive challenges. As a result, clarity as predictability may attend to facts, authorities, and inferences that the deciding court knows are wrong. This form of clarity poses its own risk—namely, that courts will ratchet up the standard for predictability ever higher, in search of predictive perfection. An example is qualified immunity. Once we see that clarity doctrines can focus on either certainty or predictability (or some combination), many opportunities for doctrinal reform and recalibration present themselves.

After establishing a clarity doctrine’s goals and basic form, courts (or other actors) can establish thresholds for finding clarity.18 Case law on how to draw clarity thresholds is surprisingly scarce, and commentators often use shorthand references to percentile confidence levels.19 For example, someone might imagine that courts should find legal clarity when they are 90 percent confident of the right answer, but not when they are 60 percent confident. These figures provide a reassuring sense of objectivity by assuming that degrees of clarity can accurately be expressed via a linear metric.20 But even if some inputs into clarity do have linear, measurable values, the presence of legal clarity often rests on interrelated and qualitatively diverse factors. Moreover, we have already seen that legal clarity ultimately rests not on a simple observation of fact, but rather on complex judgments about the propriety of an action, namely, issuing a ruling with certain consequences. And such a judgment is not entirely expressible in the language of mathematics.21 Discussing legal confidence in terms of percentages is therefore misleading despite its convenience—which probably explains why judicial opinions almost never discuss clarity in percentile terms.

But while courts are right to eschew percentile expressions of legal confidence, they frequently err in using interchangeable terminology and standards across clarity doctrines. Because each doctrine’s goals should dictate its conception of clarity, the nature of legal clarity often does and should vary from one doctrinal area to the next. So, to the extent that clarity doctrines have similar goals, they should share similar conceptions of legal clarity. And if not, then not. When courts find clarity for purposes of Chevron or the rule of lenity, they likely would not (and should not) be saying the same thing as when they find clarity for purposes of habeas corpus or qualified immunity. In other words, clarity doctrines are and should be characterized by clarity pluralism. To find out what type of clarity pluralism is best in any given doctrinal context, we have to identify and assess the purposes of the various clarity doctrines. That task is daunting, for while the law of clarity stretches far and wide, it is also threadbare.

That said, having a clearer view of legal clarity may cause us to dislike it. Clarity doctrines govern complex actions that involve numerous considerations and so are themselves often quite unclear. And we have also seen that courts might have to switch between different notions of clarity in different legal areas. What’s more, clarity doctrines can compound rather than reduce first-order disagreements about the law and also place courts in the awkward position of at least appearing to evaluate one another’s competence. Perhaps courts could mitigate these problems, such as by making principles of clarity more uniform across various doctrines or by adopting clarity rules as opposed to clarity standards. But the challenges that clarity doctrines pose could also be taken as reasons to avoid doctrinal use of clarity altogether. Courts might simply abandon many clarity doctrines, as Justice Kavanaugh has suggested,22 or replace them with more administrable procedural requirements, like voting rules.23 To evaluate those proposals, however, we must first understand just what legal clarity is and how it is used in any given doctrine. So even reforms to eliminate legal clarity implicitly call for its clarification.

The argument that follows sets out an analytic framework, along with some important applications thereof, in the hope of pointing the way toward potential reforms of various clarity doctrines. Part I outlines a theory of legal clarity by exploring its nature and the basic reasons for caring about it. As a first cut, courts should distinguish between reasons for attending to legal clarity that are grounded in certainty or predictability (or both). Each of those reasons tends toward a different “model form” of legal clarity, complete with its own distinctive strengths and difficulties. Building on those theoretical lessons, Part II explores several particularly salient and now-controversial clarity doctrines—namely, deference to state court rulings in federal habeas, Chevron, qualified immunity, the rule of lenity, constitutional avoidance, the plain meaning rule, and the good-faith exception to the exclusionary rule—to illuminate possible avenues for legal change.24 There are of course many other interesting clarity doctrines besides the ones canvassed here, including doctrines related to mandamus, injunctions, plain error, and attorney sanctions. But those other applications are less controversial and at any rate will have to await another day. Finally, Part III acknowledges and discusses ways of mitigating several possible anxieties about clarity doctrines, including worries that major clarity doctrines are too diverse, too malleable, and too awkward. Throughout, the goal is to achieve more desirable legal outcomes by clarifying clarity doctrines.

I. Analysis

This Part begins by arguing that legal clarity is best understood as a normatively grounded characterization rather than a purely empirical fact. The Part then addresses two basic reasons for attending to legal clarity: certainty and predictability. These tasks establish the analytic foundation for doctrine-specific exploration of legal clarity.

A. The Nature of Legal Clarity

The first step toward clarifying legal clarity is to recognize that it is not an empirical or linguistic fact but rather a legal characterization that ultimately rests on normative premises. Courts and even commentators sometimes elide that fundamental point, such as by conflating legal clarity with linguistic ambiguity.

An interpretation of the Constitution, statute, regulation, or precedent is often said to be clear when there is no relevant ambiguity25 or when there is no reasonable disagreement over the meaning of the law at issue.26 Thus, any doctrine featuring an “ambiguity” or “reasonable disagreement” test would qualify as a clarity doctrine. Yet courts’ tendency to link legal clarity with ambiguity and reasonable disagreement only forestalls the real question: What makes the answer to a legal question “ambiguous” or, alternatively, what makes disagreement “reasonable”?

On reflection, “legal clarity” is itself unclear in that it can refer to any number of related concepts. What unites different versions of legal clarity is the idea that a legal proposition can be more than just correct.27 In addition, considerations that tend to make a proposition correct can have other favorable implications for interpretation, such as making interpretation easier, more confident, or more widely agreed upon.28 Legal clarity is thus a second-order concept that builds on a first-order view of what qualifies as legally correct. Clarity doctrines take different positions as to how clarity builds on (and so differs from) correctness. And each doctrine is also free to adopt its own preferred conception of legal correctness. For example, clarity doctrines frequently adopt a particular interpreter’s view of what is legally correct.29 The upshot is that a legal proposition’s degree of clarity can and often does vary by context. A proposition that is legally “clear” when viewed through the prism of a given clarity doctrine may be unclear or even rejected as incorrect in another context.30

In giving more specific content to the idea of legal clarity, we must take care not to equate it with linguistic clarity. Courts and commentators frequently conflate the law with the language of statutes and other legal materials.31 Likewise, commentators on legal clarity often focus on linguistic or communicative ambiguity.32 But while the presence of legal clarity often depends in part on empirical information, such as linguistic clarity, the nature of legal clarity is ultimately a legal question. Before any empirical proposition can be legally dispositive, a legal principle has to make it so; and the choice of that principle is not itself reducible to an empirical claim. Any ruling that finds legal clarity is thus engaged in a social practice whose meaning and implications are governable by law. For example, a linguistic ambiguity could be resolved by way of an interpretive canon, yielding legal clarity.33 And conversely, linguistic clarity could fade before an interpretive principle that looks beyond the text at issue. Legal clarity, unlike a grandfather clock or its midnight chime, is not simply an object or event that can be observed and measured. It cannot be assessed solely based on empirical observation.

Instead, any finding of legal clarity represents a consequential choice by an actor (usually a court) within a legal practice and so requires recourse to law and its purposes. When courts choose to make legal outcomes turn on a particular analysis, they have an obligation to justify their choice. But while legal clarity is necessarily a legal characterization of the general type described above, it does not have to take any particular form. Legal clarity is what we make of it, and we should make it serve our purposes. So instead of asking first about language, we should begin by asking about the reasons for caring about legal clarity in the first place. Once those reasons are set, the law can fashion whatever form—or forms—of clarity it sees fit.

To get a better sense of the different ways that the law can define clarity, consider the widespread practice of reducing clarity states to numerical confidence levels.34 Writers distinguish, for example, between an “80–20” and a “60–40” confidence level.35 These expressions offer a quick and easy means of communicating degrees of confidence, but on reflection their significance is obscure. What do these numbers mean?

Perhaps they represent a tally of competing reasons, with a sufficiently lopsided score generating clarity. That approach would have the virtue of capturing the intuition that clarity has to do with the ease of arriving at a particular conclusion. But it is hard to know how many “points” to award competing arguments, and even a narrowly superior score sometimes seems capable of yielding a clear answer.

Alternatively, the numbers could depict the odds that an interpreter would stick to her opinion if she obtained perfect information about the relevant issue. Clarity would then represent a kind of confidence. But legal clarity sometimes turns on more than just access to information, as when interpreters disagree about what available information is legally dispositive. As a result, a legal issue can be unclear when judges differ on its proper resolution, even though all desirable information is known.

Finally, the percentile figures could signify the expected degree of agreement among qualified interpreters, if all such interpreters were informed and polled. This approach intuitively links clarity and consensus. But, as we will see, clarity is often thought to arise from the nature of a legal issue, independent of any counterfactual claims about imagined polls.

Whatever their intended meaning, percentile confidence levels are more at home in empirical work focused on measurable variables than normative judgments of the sort that legal clarity requires—which probably explains why percentile terms almost never appear in judicial opinions. And if they were incorporated into clarity doctrines, unadorned confidence figures would risk a false sense of precision and linearity, suppressing the cross-cutting interests actually at play. When courts assess clarity, they generally consider a range of considerations, not a single empirical matter; and even when clarity does turn on an empirical issue, the choice to attend to that issue is itself the product of a deeper, non-empirically grounded choice. Thus, a more transparent and candid approach to legal clarity would qualitatively express the considered judgment that is called for, including the legally recognized interests being weighed against one another.

Once clarity is viewed as a product of law and judgment, a range of possibilities opens up. The next Section offers a better sense of the available options.

B. Two Basic Forms

Courts care about legal clarity in the first place because of legally recognized goals, such as a desire to shift authority to specialized decisionmakers or to mitigate the unfairness of unexpected judicial rulings. This Section argues that those goals should also dictate the form of clarity that courts ultimately adopt. To continue the examples, a clarity doctrine that aims to empower technical experts might properly look different from one that is focused on mitigating unfair surprise.

In general, the idea of legal clarity can be implemented in two distinct ways: it can account for one’s own confidence in the correct answer (certainty) or the likelihood that other actors will reach a common answer (predictability). To a great extent, the distinction between certainty and predictability corresponds to the difference between first- and third-person perspectives on a given issue.36 These two forms of clarity often point in the same direction, and they can run together. When we easily reach a confident conclusion, we may expect that others will do the same. And, conversely, we may assume that questions we struggle over will prove similarly obscure to others.37

Yet a great deal can turn on whether clarity is framed in terms of certainty or predictability (or both). For example, the answer to a math puzzle might be “clear” in the sense that an expert can quickly solve it with certainty, but “unclear” in the sense that many people would toil to solve the puzzle without success—and so would fail to converge on a common prediction about how an expert would come out.38 Conversely, people who are uncertain or even wrong about which meeting place was actually stated in a garbled telephone message might nonetheless converge on a single, “clear” prediction about where the meeting will be, such as under the clock at Grand Central Station.39

As we will see, the distinction between certainty and predictability, while fundamental, is not the end of the story. The proper way to attend to certainty or predictability can be specified in myriad ways, such as via clarity rules or standards.40 And the two forms of clarity could also be blended with one another, or complemented with still other ways of approaching clarity doctrines. Thus, these two basic forms are neither mutually exclusive nor exhaustive. Certainty and predictability provide a foundation and basis for additional analysis, rather than a complete taxonomy.41

1. Certainty.

Legal clarity often exists based on a court’s degree of certainty about how to resolve a particular legal issue. Certainty might obtain when a court feels well-informed or that it is expert at figuring out the correct answer. A court could even believe that its own authority renders some of its views legally correct. By contrast, uncertainty may arise when the court feels ill-informed or inexpert, or when it worries that the law is indeterminate such that there is simply no right answer to be found.42 In these situations, the court’s reason for finding legal clarity or unclarity flows from the court’s own decision-making process. Thus, any legal clarity will run out whenever the court cannot adequately assure itself of the correct view of the law. Scholars have helpfully referred to this approach as an “internal” assessment of ambiguity.43 The key point here is that particular legal goals not only counsel in favor of attention to internal assessments, but also dictate how those assessments ought to be carried out.44

There are many ways of fleshing out clarity doctrines rooted in certainty, and we will eventually discuss several specific doctrines that can be understood in this way. For now, however, a few general observations are in order.

As an initial matter, legal clarity stemming from certainty will tend to be bounded by the scope of the relevant interpreter’s certainty. If the court’s certainty stems from detailed information, for instance, then clarity would be limited to legal issues resolvable based on that information. The source of the court’s certainty would also dictate when legal clarity arises and expires.45 As new information alters a court’s confidence about the right answer, legal clarity can wax and wane, even if every participant in the legal system agreed that there has only ever been a single correct answer.46 By contrast, unclarity stemming from a deficit in institutional competence could be persistent, on the assumption that expertise is harder to convey than simple information. And if the relevant uncertainty stems from the content of the law, perhaps because it is indeterminate, then only a relevant change in the law would have the effect of generating clarity.47

That leads to the critical but frequently neglected issue of how courts should assess the degree of certainty that yields a finding of legal clarity.48 No one confidence threshold is likely to be appropriate across all or even most doctrines, and degrees of confidence that suffice in everyday life might be either inadequate or unnecessary to qualify as certainty for the purpose of any given doctrine. For example, if the Supreme Court concluded that plaintiffs bringing constitutional tort actions have difficulty accumulating the kind of evidence necessary to establish Fourth Amendment violations, it might lower the standard for finding clearly established law, so that any supportive case law would allow plaintiffs to overcome qualified immunity.49 By contrast, courts that worried about whether officers could mount a defense might raise the relevant standard for finding clearly established law, making it more difficult for plaintiffs to prevail.50

Importantly, certainty can arise even when a legal issue is disputed or difficult to assess. As we will see, several doctrines come close to requiring a case on point before a matter can be deemed clear.51 The unfortunate consequence of those doctrines is that many lawyers are accustomed to equating legal clarity with the minimal demands of professional competence, as though there were no other way of giving clarity meaning. But precedent can call for a conclusion with certainty even when there is no case on point. Oftentimes, the best reading of one precedent calls for the creation of a new one. And when existing precedent alone offers only equivocal support for the creation of a new precedent, the need for the new precedent might still flow from other legally relevant factors, such as original history, accreted tradition, and adjacent legal principles.52

Clarity doctrines that are rooted in uncertainty may also reflect normative views pertaining to the risk of judicial error and its consequences. For example, an “error-minimization approach” might posit that the law should simply minimize the overall number of erroneous decisions. On that view, courts should set clarity thresholds in a way that maximizes the odds that later courts will reach the correct result in any given case (a standard) or set of cases (a rule). That effort could call for the adoption of accuracy-promoting heuristics, such as exclusionary rules that bar consideration of misleading or biasing information. Or clarity could be viewed as a product of a two-stage decision-making process: at the first stage, the court arrives at its own initial view of the matter; then, at the second stage, the court takes the views of other interpreters into account. If the other interpreters’ views are insufficiently probative to change the deciding court’s mind, then the issue is legally clear. If, however, the other interpreter’s views are sufficiently persuasive, then the issue is unclear—and deference to that interpreter might be appropriate.53

But minimizing overall risk might be objectionable in light of how risks are distributed among people and groups. Perhaps maximizing the number of correct cases overall is not the only thing that matters. Some cases have larger practical consequences than others, and a clarity doctrine could endeavor to minimize particularly harmful effects. Thus, a “risk-averse approach” might incorporate a precautionary principle: when the court would engender calamitous risks by adopting its own view, then it might be more inclined to find ambiguity.54 Alternatively, an “egalitarian approach” might distribute risks at least partly based on a maximin principle: the court might be more inclined to find ambiguity when it would harm vulnerable groups by adopting its own view of the law—even if a less adaptive approach would reduce error overall.55 Finally, we could imagine an “efficiency approach,” in which an initially uncertain court assesses the degree of effort that would be required to reach a certain answer to a legal question. If the effort would involve costs that exceed the benefit of reaching a certain answer, the court would declare the matter unclear, issue the appropriate ruling, and move on to the next case.56

The choice among these (and other) options is inescapable—and inescapably normative. The proper threshold for legal clarity cannot be resolved based only on empirical knowledge or the lessons of epistemology.57

Grounding legal clarity in certainty sheds light on how clarity doctrines should react to the views of other interpreters. Imagine that a court has an opinion on the correct answer to a legal question, but another interpreter disagrees. How the deciding court reacts should depend on how it understands itself to differ from the other interpreter. The views of others will matter more if the deciding court is relatively confident in the other interpreter’s reliability.58 If the other interpreter explained the basis for its view, then the deciding court might interrogate and test that rationale; if that proffered rationale is unpersuasive (or demonstrably wrong), the mere fact of disagreement would presumably have little or no additional weight. Context might also be important, such as if the other interpreter has a worse record of error on the relevant issue type or occupies an institutional role that is relatively likely to foster biases, as compared with the deciding court.59 And the existence of disagreement could itself inform a deciding court’s confidence in the other interpreter’s reliability: if the deciding court is certain about its conclusion, then the fact of disagreement might be taken as evidence that the other interpreter is unreliable.60

Viewing clarity as a product of certainty does create a distinctive risk—namely, that interpreters tend to treat their own views as not just correct but clearly correct.61 That tendency threatens to undermine the difference between clarity and mere correctness; in addition, it tends to reproduce first-order disagreements about how to decide cases, thereby generating meta-disagreement about what clarity means. To a great extent, these problems are an inescapable product of the general tendency for people to exhibit overconfidence regarding the accuracy of their own views, particularly when addressing challenging or novel questions.62 As one important study suggests, people exhibit greater bias when questions are framed in terms of internally assessed ambiguity, which is the natural frame for certainty-based clarity.63 But aspects of the legal system can exacerbate overconfidence bias. On balance, a confident court might seem more legitimate. Judges accordingly write opinions in the style of briefs that contest every point, admitting no weakness. And the high stakes in many cases can make uncertainty difficult to admit. Alas, we will see that key doctrines rooted in certainty sometimes invite these problems by failing to explain how judges should go about identifying views that are certain, as opposed to merely correct.64

2. Predictability.

The second basic reason for incorporating legal clarity into the law has to do with how predictably certain interpreters will decide cases.65 Courts might want to make allowances for other actors—whether they be private parties, executive officials, or even judges—who have had to make difficult predictions about legal outcomes.66 A clarity doctrine might also promote fairness by preventing people from being held responsible based on rules that were not apparent when they acted.67 Finally, unpredictable rulings can be disruptive, yielding institutional and social costs that clarity doctrines might mitigate.68 In all these cases, attention to predictability privileges the perspective of other decisionmakers, apart from the deciding court. Thus, a court could attempt to assess and take account of the views of other actors, even when the court itself is “internally” certain that the other actor’s reasonable view is incorrect. Scholars have labeled this basic approach an “external” assessment of ambiguity, by which one interpreter attempts to predict or imagine how other interpreters would resolve a particular issue.69

Clarity doctrines can account for predictability in several different ways, as we will see below. For now, it is enough to outline the key implementation questions, most of which have to do with the vantage point from which predictability is assessed.

When considering predictability, the most basic issue is: Whose perspective matters?70 The answer depends on the law’s goals. The most obvious reasons to care about predictability have to do with the burdens it places on decisionmakers. For that reason, doctrines are often sensitive to predictability when decisionmakers experience severe harm from a “bad bet” on what the law is. The plight of a criminal defendant surprised by the law’s reach offers a pat example.71 However, courts might attend to another decisionmaker’s perspective even if the decisionmaker herself does not experience any harm from bad bets. Even if the decisionmaker is not an object of sympathy—or insufficiently so—her need to grapple with unpredictability could still objectionably harm third parties. The risk that unforeseeable civil liability might deter officials from undertaking socially beneficial actions, for instance, is a leading basis for qualified immunity.72

Besides knowing which decisionmaker matters, we also need to know what information that actor possesses and what analysis she can or should undertake. An expert steeped in a particular field might be confident of how the Supreme Court would rule on a given issue, but that expertise would be inaccessible to most people who grapple with legal questions on a daily basis. Attention to predictability accordingly counsels in favor of privileging certain bases for prediction while discounting or excluding other sources of admittedly useful information. Some doctrines privilege case law, or case law from certain courts, over relatively inaccessible or disputable factors like history, structure, or policy.73 As we will see, a number of doctrines also focus on extant precedent, rather than calling for an all-things-considered evaluation of how certain courts would rule.74 Similar reasoning might privilege statutory text over relatively inaccessible sources of legislative history or atextual legislative purposes.75

Temporality is also critical to predictability. If a businessperson, police officer, or court must decide at Time 1, it would usually make little sense to ascertain the relevant law’s predictability at Time 2. Instead, predictability is most naturally assessed when the relevant actor actually decided.76 But not always. A doctrine might hope to encourage decision-making at the right time and so penalize decisions that were made at the wrong time or in the wrong way. When there is no need for haste, for example, a clarity doctrine might assess unpredictability at the point in time after the decisionmaker actually acted, so as to penalize decisionmakers who chose to act when only imperfect information was available.77 Issues of temporality thus bleed into questions about how to create incentives for optimal information access: a greater delay might have allowed additional research or reflection, thereby enhancing predictability.

A clarity doctrine that is rooted in predictability will also tend to generate different clarity thresholds in different contexts. After all, each clarity doctrine is concerned with a different decisionmaker operating under different conditions and with different obligations. What matters is the relevant decisionmaker’s actual and prescribed process for arriving at legal conclusions. And that process could diverge from the deciding court’s analogous process. For example, a deciding court that aims to protect potential criminal defendants from the costs of “guessing wrong” about the law might assess predictability from the standpoint of someone without access to an attorney.78 But when that same court turns to resolve the relevant issue, it would presumably be prepared to use the full range of interpretive resources and methods available to it.

There is another reason why doctrines rooted in predictability will tend to have varying clarity thresholds: these doctrines could very well be concerned with more than just the effects of unpredictability. Two interrelated considerations are particularly likely to press against the goal of fostering predictive clarity. First, even after allowing for the importance of fostering predictability, the legal system is still concerned with enforcing the law as correctly understood. And demanding perfect certainty before applying the law would effectively repeal the law, or disable lawmaking authority, since virtually any new scenario raises interpretive questions or is separable from precedent in some way.79 That kind of worry pushes against making too many allowances for unpredictability. Second, the legal system will often want to encourage decisionmakers to act as correctly as they can, even when struggling with uncertainty. That incentives-based consideration counsels against making allowances for unpredictability when doing so will reduce the cost of error.

Grounding legal clarity in predictability also sheds light on how courts should react to views of other interpreters—and does so in a way that is qualitatively different from cases when legal clarity rests on certainty. A descriptive approach (focusing on what the other interpreter would do) is appropriate for a clarity doctrine that aspires to mitigate the disruptive effect that a surprising legal ruling would in fact cause, whereas a prescriptive approach (focusing on what the other interpreter should do) is more appropriate for clarity doctrines that aim to encourage appropriate interpretive efforts. For example, a court might adopt a descriptively or prescriptively grounded expectation that local officials will stay especially apprised of regional precedents, as opposed to precedents from other jurisdictions.80 In this situation, a deciding court could very well conclude both that it is sure of the correct answer and that the local official would or should have arrived at a different result based on local guidance. Legal unpredictability would then exist even when the deciding court is perfectly certain as to the correct result.

What’s more, a court concerned with predictability often has good reason to be interested in the views of interpreters who are discernably wrong. For example, a court concerned with predictability must think about whatever authorities—such as extant precedents—are available to the actor who is engaged in prediction. By contrast, a judge confronting uncertainty would have reason to focus exclusively on the views of epistemic “peers” or “friends,” since those actors would better help the deciding court understand the correct answer to the question at hand.81 Thus, the goals of each clarity doctrine in effect dictate the relevant scope of legal authority. Doctrines concerned with certainty might pay limited attention to disagreements in which one side has demonstrably erred. By contrast, doctrines focused on predictability should often care about disagreement in itself, even when one side is wrong, given that some actors must or should make predictions without the luxury of disqualifying authoritative views that turn out to be wrong.

Predictability does come with its own characteristic risk, one that is almost the mirror image of the risk posed by certainty. Once interpreters set predictability as their lodestar, they may be tempted to raise the standard for clarity ever higher. What starts out as a rule that most qualified interpreters must agree can turn into a requirement that clarity exists only if there is a specific case or statute precisely on point. This tendency stems in part from the understandable desire for objective evidence of predictability, as opposed to hypothetical assertions about what most or all relevant interpreters would do. After all, only with that kind of objective evidence can a predictability-based inquiry itself be fully predictable. But there is another important reason to ratchet up the standard for clarity: when courts use predictability to find clarity, they could be viewed as making a negative comment on any interpreters who came out the other way. And, as we will see,82 courts have good reason to pause before casting aspersions on judges as well as other interpreters. That reluctance can create an incentive to avoid finding predictive clarity at all.

* * *

This Part has emphasized a basic choice underlying the construction of clarity doctrines—namely, the choice to attend to either certainty or predictability (or both). Because clarity doctrines govern complex judgments involving conflicting considerations, there can be no ironclad rules about how any given doctrine either will or should be implemented. However, the logic underlying the distinction between certainty and predictability points toward two model forms of legal clarity, as shown below.

Table 1: Two Model Forms of Legal Clarity

Clarity as Certainty	Clarity as Predictability
Perspective of the deciding court	Perspective of an actor other than the deciding court
All-things-considered judgments, or else accuracy-promoting heuristics	Judgments based on limited information available to another actor
Temporally grounded in the present—that is, in the moment of the deciding court’s ruling	Temporally grounded in the moment of the other actor’s decision
Accounts for other interpreters’ superior expertise and/or privileges the views of like-minded interpreters over opposing ones	Accounts for the existence of disagreement among relevant authorities, even if some of the authorities are discernibly incorrect
Risk of collapsing the clarity inquiry into the underlying merits inquiry	Risk of escalating the clarity standard and demanding specific law on point

In general, focusing on certainty makes the most sense when accuracy is at a premium and the deciding court is especially adept at gleaning the correct view of the law. By contrast, focusing on predictability is generally attractive when a doctrine aims either to make allowances for the interpretive challenges faced by privileged actors or to ensure merely competent decision-making by less expert interpreters. With those foundational points in mind, it is time to turn to more specific applications.

II. Applications

The law establishes various forms of legal clarity, but that pluralism is often elided and underappreciated.83 This Part explores several well-known doctrines that turn on the presence or absence of legal clarity. Though courts use similar terminology in all these areas, each of these doctrines is shaped by its distinctive, legally recognized purposes—particularly the perceived need to focus on either certainty or predictability. That basic point opens up opportunities to question each doctrine’s accepted purposes and thereby support reforms, including changes in clarity thresholds, the admissible sources of clear law, and the implications of disagreement among courts.

A. Section 2254(d)(1)

Let me begin with a legal doctrine that featured a judicial debate, however muddled, between two starkly contrasting visions of legal clarity—one focused on certainty and the other on predictability.

Under what is often called § 2254(d)(1), a statutory provision amended by the Anti-Terrorism and Effective Death Penalty Act of 199684 (AEDPA), federal habeas courts may grant relief when a state court’s adjudication on the merits has “resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.”85 These words establish a clarity doctrine: they not only contemplate a distinction between “unreasonable” and reasonable applications of law, but also demand that the applied law be “clearly,” as opposed to ambiguously, “established.” In the seminal case Terry Williams v Taylor,86 the Court split 5–4 over the proper interpretation of this provision.87 Everything the Court has subsequently decided about § 2254(d)(1) is derivative of the picture of legal clarity adopted in Terry Williams. Yet the justices’ opinions obscured the nature of the Court’s chosen approach, as well as the costs of adopting it, as opposed to other legally viable alternatives.

Writing for the Court, Justice Sandra Day O’Connor advocated an “objective” approach to unreasonableness that relied on Court precedent “as of the time of the relevant state-court decision.”88 That analytic framework focused not on whether the state court reached the wrong outcome, all things considered, but rather on whether the state court issued a competent ruling given the “established” Court precedent then available. Elaborating that point, the Court has subsequently explained that relief is available when the state court has committed “an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement.”89 That test emphasizes predictability: surely many legal propositions can be known with a high degree of certainty, even if they are not “well understood and comprehended in existing law”—here, Court precedent—at the time that a state court issues a particular ruling.

By contrast, Justice John Paul Stevens’s four-justice opinion on this issue advanced a more certainty-oriented approach to § 2254(d)(1).90 Here is Justice Stevens’s most succinct statement of how he would decide whether a state court determination on the merits amounted to an “unreasonable application of[ ] clearly established” Court precedent:

Our difference is as to the cases in which, at first blush, a state-court judgment seems entirely reasonable, but thorough analysis by a federal court produces a firm conviction that that judgment is infected by constitutional error. In our view, such an erroneous judgment is “unreasonable” within the meaning of the Act even though that conclusion was not immediately apparent.91

Justice Stevens is describing the various confidence states that can arise within a regime focused on certainty-based clarity. The federal court thinks “at first blush” that the state court has been “entirely reasonable,” but “thorough analysis” later generates not just a belief but a “firm conviction” of error. In other words, the state court ruling should be upheld as reasonable under § 2254(d)(1) when the federal court suspects or believes—if it merely “seems”—that an error occurred. By contrast, habeas relief is available whenever a federal court arrives at a “firm conviction” that an error took place.92

Unfortunately, neither side of this dispute directly defended its preference for resting legal clarity on certainty or predictability. Justice O’Connor’s majority opinion acknowledged that “[t]he term ‘unreasonable’ is no doubt difficult to define,” but found solace in the fact that “it is a common term in the legal world and, accordingly, federal judges are familiar with its meaning.”93 This unilluminating observation reflects courts’ tendency to cross-apply terminology in different doctrinal areas. Justice O’Connor added an additional, rather tautological claim: “For purposes of today’s opinion, the most important point is that an unreasonable application of federal law is different from an incorrect application of federal law.”94 Justice O’Connor appeared to view that point as significant on the assumption that Justice Stevens would in effect apply a de novo standard of review, thereby effacing the statute’s reference to “unreasonable” state court decisions. As we have seen, however, Justice Stevens focused on certainty and so (charitably read95 ) was simply advancing a different way to ascertain reasonableness.96 The key unanswered question, then, is whether the Court’s focus on predictability was warranted.

To answer that question and justify the Court’s predictability-based approach, one would have to identify the legally accepted purposes of federal habeas corpus. According to the Court, federal habeas relief is not justified whenever it would vindicate an individual’s rights. On the contrary, states’ interests in finality are thought to defeat the interests of habeas claimants who have had a full and fair opportunity to raise their claims on direct appeal, even if the claimants have suffered rights violations.97 Given that view, federal habeas relief is justified only if supported by structural interests, over and above the interest in vindicating rights. When state courts don’t just err in resolving open questions but actually defy then-extant Supreme Court precedents, they cast into doubt what was previously settled and so degrade the overall supremacy and uniformity of federal law. The doctrine accordingly posits that the real point of federal habeas is to adjudge whether the state is incompetent or disobedient—an outlook that has the benefit of explaining why § 2254(d) expressly focuses on precedents issued by the Supreme Court, the only federal court that state courts are bound to follow.98 And, years after Terry Williams, the Court eventually came close to saying as much: “Section 2254(d) reflects the view that habeas corpus is a ‘guard against extreme malfunctions in the state criminal justice systems,’ not a substitute for ordinary error correction through appeal.”99 That thinly explained view of federal habeas corpus—that it checks only “extreme malfunctions” by state courts—implicitly underwrites the version of legal clarity operative in the Court’s extensive case law applying § 2254(d)(1).

Justice Stevens defended his focus on certainty by offering a different account of what federal habeas is for: vindicating rights.100 But Justice Stevens never explained why his approach wouldn’t lead to de novo review, thereby eliding the statute’s “unreasonableness” requirement in precisely the way that Justice O’Connor alleged.101 A certainty-based approach to legal clarity can offer a solution to this problem: when a federal court disagrees with the state court decision under review but nonetheless lacks a high degree of confidence, a certainty-based approach would call for deference to the state court. Put differently, deference to state court adjudications might be proper only when, in light of the state court’s view, the federal court is significantly unsure about the correct view of the law.102

To see how attention to certainty can yield greater protection to individual rights even as it gives genuine effect to AEDPA’s “clearly established” law requirement, consider a pair of related rulings. In Yarborough v Alvarado,103 the Court denied that its case law had clearly answered whether a minor detainee’s youth should figure into a Miranda custody analysis.104 Later, in J.D.B. v North Carolina,105 the Court answered the underlying merits question with a confident yes.106 Both cases came down 5–4, with Justice Anthony Kennedy casting the key “fifth” vote in each ruling. Kennedy’s votes make sense under a predictive approach to § 2254(d)(1). As his majority opinion in Alvarado emphasized, “Our Court has not stated that a suspect’s age or experience is relevant to the Miranda custody analysis.”107 Justice Kennedy’s decision to find unclarity in Alvarado based on gaps in what the court had expressly “stated” reflected a sensible concern for predictability. But as J.D.B. would later confirm, extant Court precedent did already afford a strong, even if implicit, basis for considering a minor detainee’s age.108 Thus, the Alvarado dissenters (including Justice Stevens) also adhered to a sensible conception of legal clarity. In finding a violation of clearly established Supreme Court precedent, the Alvarado dissenters followed a certainty-oriented approach to legal clarity and anticipated the confident merits reasoning that later prevailed in J.D.B.109

There are several other, more limited ways to flesh out Justice Stevens’s certainty-based approach to federal habeas review of state court adjudications. One straightforward possibility is that federal habeas courts should bow to state-court interpretations that take advantage of otherwise underappreciated epistemic advantages. For instance, state courts could have superior access to critical facts by virtue of having direct access to hearings, trials, and local experience.110 On that approach, § 2254(d)(1) would still preserve room for habeas relief when the federal court is certain, on the cold record, that an individual’s rights had been infringed. Alternatively, federal habeas courts might have to consider whether their interventions in close cases would vindicate a sufficiently weighty constitutional interest as to outweigh competing interests in finality and efficiency.111 Justice Stevens might even have had that last possibility in mind when he wrote that relief under § 2254(d)(1) requires “concluding that [state-court decisions] were infected by constitutional error sufficiently serious to warrant the issuance of the writ.”112

The difference between certainty and predictability also points toward qualitatively different clarity thresholds. Again, courts should find legal clarity when doing so would achieve the relevant doctrine’s goals. In part because of its focus on predictability, the Court has adopted a stringent approach when assessing reasonableness under § 2254(d)(1), in many cases (including Alvarado) virtually demanding a specific Supreme Court holding that indisputably governs on the precise facts at hand.113 “Even ‘clear error’ will not suffice.”114 That approach makes sense only on the assumption that predictability concerns predominate over concerns about certainty regarding the right legal outcome. By contrast, a proponent of a certainty-based approach would allow federal courts to afford relief when they are sufficiently confident that the state court has erred. Under that alternative inquiry, § 2254(d)(1) might be sensitive to the strength of the briefing at hand, as well as to the federal court’s familiarity with the relevant issues.115

Finally, the Court’s emphasis on predictability also informs how § 2254(d)(1) responds to the views of lower courts. This issue yielded an interesting point of accord between the dueling opinions in Terry Williams. Justice O’Connor denied that the appropriate question was whether “the state court has applied federal law ‘in a manner that reasonable jurists would all agree is unreasonable.’”116 In her view, a federal court should ask about “objective unreasonableness,” and so should not rest “its determination [ ] on the simple fact that at least one of the Nation’s jurists has applied the relevant federal law in the same manner the state court did in the habeas petitioner’s case.”117 Meanwhile, Justice Stevens likewise denied that a single reasonable jurist “should always have greater weight than the contrary, considered judgment of several other reasonable judges.”118 In one sense, this concordance is sensible: the fact that courts disagree does not in itself establish either uncertainty or unpredictability. A limited judicial disagreement could conceivably result from a single, unforeseeable blunder. So, if the Terry Williams opinions meant to make only that modest point, they were on solid ground.

Yet we have already seen that disagreement should often inform judgments regarding both certainty and predictability, albeit in qualitatively different ways. And, in fact, the Court has come close to recognizing as much. For example, the Court has recently suggested that when “lower courts have diverged widely” in reaching a particular question, that state of affairs reflects a “lack of guidance” from the justices.119 That view is in tension with Terry Williams’s rejection of the “all reasonable jurists” test, but it makes good sense—at least if we follow the Court in focusing on predictability. The more often that either state or federal courts reach the same result as the state court under review, the harder it is to say that the legal issue had a predictable contrary answer, that the state court exhibited incompetence, or that the state court’s ruling eroded guidance in existing Court precedent. And as we have seen, emphasizing that sort of objective evidence has additional appeal because it promotes predictability in how the clarity doctrine itself is applied.

Again, greater attention to certainty would support a different approach. Lower court dissensus is less likely to undermine the justices’ certainty regarding their own conclusions.120 The justices, after all, do not typically view relatively resource-strapped lower courts as their epistemic peers.121 And the justices are often both willing and able to identify what they regard as errors in lower-court opinions. Thus, extant disagreement is less likely to generate unclarity under a certainty-based approach.

Finally, a certainty-based approach would help defend and recast the circuit courts’ presently forbidden practice of drawing on their own case law when identifying clearly established Supreme Court precedent, including when the circuit rulings postdate the state court adjudication at issue.122 Circuit judges can plausibly view their peers’ applications of Court precedent as evidence of what that precedent had previously established. Thus, circuit precedent can help persuade a circuit court that it has correctly read the Court’s case law—and that the state court has certainly erred.123 Yet the Court has plausible reasons for resisting that approach, given its strong emphasis on prediction. After all, predictions are most reliable when there is on-point Court precedent, mooting any need to sift through general principles or circuit case law.

In sum, several aspects of the Court’s § 2254(d)(1) jurisprudence, including its eventual willingness to place considerable stock in circuit splits, is the natural result of its long-time focus on predictability, starting with Terry Williams. And the Court’s focus on predictability in turn depends on its disputable view of the core purposes of federal habeas. Reformers who would celebrate a more claimant-friendly approach to federal habeas should accordingly challenge the Court’s account of federal habeas and thereby propose a more certainty-oriented approach to clarity. Doing so would redeem the unfulfilled analytical promise in Justice Stevens’s Terry Williams opinion.

B. Chevron

Chevron U.S.A. Inc v Natural Resources Defense Council, Inc124 announced that administrative agencies are often entitled to deference regarding the meaning of the statutes they administer.125 To mark the limits of that principle, courts ask whether “the intent of Congress is clear . . . for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress.”126 And the Court further explained that “a court may not substitute its own construction of a statutory provision for a reasonable interpretation made by the administrator of an agency.”127 In short, an agency can prevail even if the court disagrees with the agency’s reading of the relevant statute—but only if the statute is unclear.

Chevron’s clarity rule is primarily concerned with certainty.128 Whereas clarity doctrines focused on predictability are typically explicit in identifying another interpreter whose perspective must be taken in account, Chevron speaks of whether legislative intent is “clear” or “unambiguous,” full stop. That formulation suggests that the relevant perspective is the reviewing court’s—unmediated by any expectations of how any other decisionmaker would behave. Other features of the Chevron ruling also reflect a certainty-based approach. To wit, Chevron notes: “If a court, employing traditional tools of statutory construction, ascertains that Congress had an intention on the precise question at issue, that intention is the law and must be given effect.”129 The implication is that ambiguity arises when the “court” itself “ascertains” no legislative intent—in other words, when the court cannot glean the correct legal answer.130

Chevron practice also suggests that clarity corresponds to certainty and ambiguity to uncertainty. Chevron cases do not hold both that the meaning of a statute is X and that an agency reasonably concluded that the meaning is Y. Instead, Chevron decisions typically rule either that the agency’s view is unreasonable, in which case the court is certain that the agency is wrong, or that the agency’s view is reasonable, such that the agency’s view (it seems) may actually be correct.131 By contrast, jurisprudences grounded in predictability—such as qualified immunity, discussed below—do sometimes yield confident holdings on the merits while simultaneously declaring contrary views to be reasonable.132 That combination of outcomes makes sense when clarity rests on predictability, since the point of focusing on unpredictability is to attend to another’s point of view, even if the deciding entity itself is certain that it is right. When certainty-based clarity is at play, by contrast, the Court cannot both have a sure view of the correct answer and still find a contrary view to be reasonable.

The two dominant theories of Chevron respectively rest on notions of implied delegations and relative expertise, and both of those theories support the doctrine’s attention to certainty. First, the implied delegation theory assumes, as a legal fiction, that Congress intentionally creates statutory ambiguities in order to leave pertinent gap-filling decisions to agencies.133 On that theory, a court that cannot ascertain Congress’s intent with adequate confidence should infer that a delegation has taken place—and presume that the agency’s view is reasonable. By contrast, a court that can confidently ascertain Congress’s intent should declare as much, even if doing so means invalidating the agency’s view. Second, the comparative expertise theory maintains that specialist agencies know more about their own statutes than do generalist courts.134 Therefore, courts that lack strong confidence in the meaning of the statutes should defer to the relevant agency’s views on epistemic grounds. Here, too, a court can and should override the agency’s expert interpretation, provided that the court is sufficiently confident it is correct.

Chevron’s focus on certainty bears on its susceptibility to uniform application across judges. Because criteria for legal correctness often vary by judge, so too do the conditions for finding certainty-based clarity.135 In the Chevron context, different judges read statutes with different methods in mind and so often have varying levels of confidence about questions of statutory interpretation. A textualist might believe that recourse to a single statutory text is nearly conclusive evidence of statutory meaning, whereas a non-textualist might additionally seek out legislative history and ponder legislative purposes. When text points strongly in one direction and other evidence points in another, the textualist might have higher confidence than her non-textualist colleague. It is also possible for a text to be cryptic while non-textual evidence points strongly in a particular direction, in which case the non-textualist might have a higher level of confidence.136 Which scenario predominates is ultimately an empirical question, but it is plausible to think that attention to more sources of evidence at least sometimes tends to increase the odds of finding conflicting evidence and, therefore, of uncertainty.137

The relationship between legal methodology and certainty should also inform Chevron’s clarity threshold. As we have just seen, textualists exclude many potentially conflicting inputs into legal meaning and so may more readily reach high levels of confidence regarding the meaning of any given statutory text. The best reading of the text in isolation, in other words, might simply drown out all other factors. Further, textualists have less expertise-based need to defer to agencies because courts and agencies have relatively comparable expertise when it comes to understanding the plain meaning of statutory texts (as opposed to, say, the meaning of scientific studies). Thus, text-oriented jurists may find legal certainty—and clarity for purposes of Chevron—more often or more readily than their less textualist colleagues. And that is just what many jurists have observed, including Justices Antonin Scalia, Elena Kagan, and Brett Kavanaugh. "</a

More generally, judges’ views of legal correctness affect how they set clarity thresholds. Again, knowing a judge’s level of confidence is insufficient to know whether there is clarity under any given doctrine, for different doctrines can set clarity thresholds with varying degrees of stringency or quality. To illustrate the range of possible clarity thresholds, imagine a judge who is somewhat confident that Congress intended X. If the judge subscribed to an implied delegation theory of Chevron, the judge would have to describe the kinds of statutory clarity that indicate a legislative decision to resolve an issue, rather than delegating it to an agency. If some issues are especially unlikely to be objects of delegation, for instance, then the required degree of confidence might rise.139 By contrast, if the judge subscribed to an expertise-based view of Chevron, she might conclude that agency readings should override the judge’s own views whenever the fact of an agency’s authoritative reading persuades the judge to change her mind; and the judge might then assess how much persuasive force to give an agency’s authoritative view.140 Of course, each step in the foregoing reasoning would require justification—and varying any of those steps would give rise to a new version of Chevron.

So when judges of different methodological stripes faithfully aspire to implement Chevron’s famous test, they will end up differing on what qualifies as a “clear” interpretation or the “unambiguously expressed intent of Congress.”141 Far from purporting to resolve the correct means of reading a statute, Chevron reproduces preexisting disputes over method by simply calling for use of “the traditional tools of statutory construction.”142 So when there is disagreement on just what those tools are and how to use them, disagreement about Chevron will follow. And even when judges concur on the type of certainty that creates legal clarity, they may have different views of how or where to mark the clarity threshold, depending on how they understand the goals of Chevron deference.143

In a recent paper, then-Judge Kavanaugh drew attention to the challenges embedded in clarity doctrines, particularly Chevron.144 In Justice Kavanaugh’s view, clarity doctrines generally involve two separate and intractable problems:

First, judges must decide how much clarity is needed to call a statute clear. If the statute is 60-40 in one direction, is that enough to call it clear? How about 80-20? Who knows?

Second, let’s imagine that we could agree on an 80-20 clarity threshold. In other words, suppose that judges may call a text “clear” only if it is 80-20 or more clear in one direction. Even if we say that 80-20 is the necessary level of clear, how do we then apply that 80-20 formula to particular statutory text? Again, who knows? Determining the level of ambiguity in a given piece of statutory language is often not possible in any rational way.145

In essence, Justice Kavanaugh is saying that the law affords no principled basis for choosing between 60–40 and 80–20 as clarity thresholds and that, even if it did, we still wouldn’t know when those numerical conditions are satisfied. Because he believes that “there is no neutral method to evaluate whether a text is clear or ambiguous,” Justice Kavanaugh concludes that clarity findings represent “a certain sort of ipse dixit”—and are a threat to neutral adjudication.146 Justice Kavanaugh therefore recommends that many doctrines, including Chevron, be either scaled back or abandoned altogether.147

Justice Kavanaugh is right that extant case law presently affords precious little guidance on how to go about making a clarity finding, particularly when it comes to ascertaining clarity thresholds. In that respect, he anticipates one of the basic claims of this Article—namely, that there is much to be gained by clarifying the nature of legal clarity. By now, however, we should be unsurprised by Justice Kavanaugh’s observation that the Chevron standard is variable across judges. Whenever a clarity doctrine is rooted in uncertainty, its implications will depend in part on how decisionmakers approach underlying merits issues. In eliding that point and tacitly treating ambiguity as a singular concept, current doctrine indeed fosters confusion.

Yet Justice Kavanaugh is wrong to suggest that the search for a principled approach to legal clarity is futile or, as he puts it, “not possible in any rational way.”148 There are many plausible ways of understanding legal clarity, depending on the accepted goals of the underlying doctrine. If the Supreme Court so desired, it could authoritatively establish a particular approach. For example, we have seen that the Court could begin to set a more robust clarity threshold by directing judges to side with agency readings whenever the agency’s relative expertise is sufficiently compelling to persuade the court to change its view of legislative intent—a certainty-based approach that would parallel Justice Stevens’s proposed approach to § 2254(d)(1), discussed above.149 Providing that kind of guidance would not eliminate judicial disagreement—as if such a thing were possible or desirable—but it would foster a “rational” method.

True, Chevron’s clarity threshold still wouldn’t be reducible to a percentile confidence level akin to Justice Kavanaugh’s 60–40 expression, but we have also seen that these expressions are misleading and that any given clarity threshold is defensible only to the extent that it achieves legally recognized goals.150 In the Chevron context, salient potential goals include improving accuracy in outcomes, encouraging efficient legislative drafting, and fostering democratic influence over agency decision-making.151 And once the goals are identified, they can be implemented in a more or less rule-like way. These aspects of Chevron’s doctrinal implementation are unexceptional, as almost all legal principles rely on qualitative judgements, rather than quantitative ratios. The need to incorporate judgments into clarity thresholds might supply reason to change or specify the law of clarity, but that conclusion would not mean that law has come to an end. Clarity thresholds are not any less “neutral” than other areas of judicial decision-making that call for judgment, such as what process is “due” or what treatment is “equal.”152 Surely the law can tolerate some inevitable indeterminacy about the meaning of clarity.

The analysis thus far also points toward another response to worries about evenhanded administration: turning away from certainty and toward predictability. A court could ask, not what readings are reasonable from its perspective, but rather what a reasonable agency could have concluded about the relevant statute. That approach would also accord with some plausible accounts of agency deference. Perhaps Congress writes not for courts to read statutes according to their own lights, but rather to give agencies space to employ policy expertise, so long as they also exhibit legal competence.153 Chevron would then cast agencies somewhat in the same position as state courts under current interpretations of § 2254(d)(1).154 Like state courts, the agency would be expected to be familiar with relevant case law, including the various judicial methods of statutory interpretation. But, also like state courts, the agency would prompt federal court intervention only when it has failed to anticipate foreseeable rulings.155 On that recasting, Chevron would be more concerned with judicial disagreement, as evidenced by circuit splits156 and dissents among the justices. This response, too, would be rational and feasible. And it would plausibly advance Chevron’s long-accepted goals while reducing inter-judge disparities.

Finally, the existence of interpretive disagreement might have different implications depending on whether Chevron is viewed as resting on certainty or predictability. Insofar as Chevron attends to certainty, the mere fact of disagreement could matter, but only when jurists view other, disagreeing colleagues as comparably effective and knowledgeable interpreters.157 Given the inconsistent quality of briefing and resources at different levels of the judicial hierarchy, judges are most likely to be justified in viewing one another as epistemic peers when they occupy the same level of the judicial hierarchy and have access to information of substantially the same quality.158 The paradigmatic case would arise when a justice sees dissensus among her colleagues. But even then, a justice who attends to uncertainty might greatly discount the disagreement of others for various reasons, including methodological disagreements or simply a high level of confidence in her own view.159 It is after all commonplace—and quite rational—for one judge to discount another’s view when that view expressly rests on what seems to be a mistaken fact or inference.160 For all these reasons, it is unsurprising that current Chevron practice, which attends to certainty, does not afford circuit splits and other forms of judicial disagreement a significant role.

But the fact of disagreement would take on a very different appearance if predictability entered into Chevron doctrine. No agency, after all, would have the luxury of ignoring circuit court rulings or discounting mainstream interpretive methodologies like textualism or purposivism. Thus, the views of other courts would play a larger role in a predictability-based Chevron jurisprudence. Moreover, the precise way of accounting for those other views would not flow from generally applicable epistemic premises, as some commentators have assumed, but rather from the doctrine’s legally recognized goals.161 As a result, Chevron doctrine could be justified in attending to the legal opinions of various groups. If the goal is to foster predictability across the federal system, for example, then perhaps Chevron should equally account for the views of all federal judges, including rulings by circuit court judges. Alternatively, continuing concern for certainty could warrant special or exclusive attention to jurists who qualify as epistemic peers.162 In any event, attention to predictability would coexist with other goals and so would not afford dispositive weight to the mere fact of disagreement among federal judges, or even among the justices themselves.163 For one thing, a requirement of near judicial unanimity would place heavy burdens on legislative drafters to speak with care, lest a single outlier judge rule erroneously and thereby imbue an agency with discretionary regulatory power. Concerns about defeating legislative intent and administrative flexibility would thus counsel in favor of tempering attention to actual disagreement with a substantive requirement that an ambiguity finding must rest on at least some uncertainty regarding the correct reading.

In sum, Chevron deference is presently a clarity doctrine focused on certainty, and that approach seems justified, at least if we accept the doctrine’s classic justifications. But Chevron has lately come in for criticism on grounds that highlight the weaknesses of certainty-based clarity, and those concerns could support adoption of a more predictability-based approach. Whether Chevron deference survives its critics may depend on whether the doctrine itself becomes clearer.

C. Qualified Immunity

Qualified immunity is under fire for many reasons,164 but attention to its use of legal clarity points out a new dimension of potential critique and reform. Perhaps because it has defined qualified immunity’s purposes increasingly narrowly, the Court has neglected the possibility that certainty-based considerations could, and should, play a significant role in ascertaining the scope of clearly established case law.

Under qualified immunity doctrine, executive officials are generally immune from civil liability unless their actions transgress “clearly established statutory or constitutional rights of which a reasonable person would have known.”165 That test cashes out the idea of legal clarity in terms of reasonable disagreement: when a plaintiff’s right is “clearly established,” a “reasonable” defendant is said to be aware of them.166 By contrast, an officer and a court reasonably disagree when the law is unclear.

Like the Court’s reading of § 2254(d)(1), qualified immunity focuses on predictability.167 The Court has made clear that the qualified immunity inquiry focuses on the standpoint of the officer at the time of the relevant conduct—not on whether the deciding court can assure itself of whether the officer acted lawfully.168 In other words, the doctrine is concerned with the predictive position of an executive official.169 Confirming as much, the doctrine emphasizes the precise timing of the officer’s decisions, as well as authorities the officer would or should consult.170 In addition, the Court sometimes rules in favor of qualified immunity even as it confidently concludes that a constitutional violation has taken place.171 In those cases, the Court reasons that qualified immunity’s direct beneficiaries are situated differently than courts: “The official cannot be expected to predict the future course of constitutional law.”172 Yet the Court’s willingness to make allowances for official error has its limits. The doctrine imagines a stylized executive official who has at least some familiarity with extant case law, thereby encouraging officials and departments to assimilate new case law into training and policy.173 In this and other ways, the doctrine tempers its focus on prediction so as to promote other interests.174

Qualified immunity’s legally recognized purposes support its focus on predictability. At various times, the Court has aimed to reduce the risk of officials’ personal liability both to encourage the officials to exercise their full authority and to allow them to avoid litigation burdens that might distract from their jobs.175 These justifications all support the same goal: reducing competent officials’ liability. In declaring that “[qualified immunity] protects ‘all but the plainly incompetent or those who knowingly violate the law,’”176 the Court makes officer protection the key value, not legal correctness. At times, the Court has even come close to requiring that the officers transgressed an extant case on point.177 In that respect, too, qualified immunity parallels § 2254(d)(1).178

Yet there are good reasons why qualified immunity should also attend to certainty. Most obviously, the doctrine could account for the interests in affording compensation to civil rights claimants, as well as the need to encourage officers to attend to the Constitution.179 At an extreme, the Court might find legal clarity so rarely that many self-interested officers wouldn’t think twice before engaging in conduct that would likely be held illegal. That state of affairs would in effect negate the 42 USC § 1983 damages remedy and reduce the practical force of the underlying rights at issue. These points raise the possibility of complicating extant qualified immunity doctrine by requiring some inquiry focusing on certainty—a reform possibility that is analogous to one that we have already seen in connection with § 2254(d)(1).180

In a similar vein, Professor John Jeffries has suggested that the doctrine stop asking whether the defendant infringed a “clearly established” right and instead ask whether the relevant conduct was “clearly unconstitutional.”181 Jeffries acknowledges that, “in many circumstances, there would be no difference between the formulations.”182 Yet the proposed reformulation would matter.183 “Asking whether conduct is ‘clearly unconstitutional,’” he suggests, “is less tied to precedent and less technical.”184 “Most importantly,” Jeffries continues, “it incorporates the notion of common social duty.”185 Thus, under Jeffries’s proposal, qualified immunity would be unavailable for “outrageousness,” “egregiousness,” or “truly appalling” conduct “whether or not the specific misconduct has been adjudicated before.”186 However, even that aspect of Jeffries’s proposed test seems to echo current doctrine, which likewise recognizes a rare set of extreme constitutional violations that are obviously unconstitutional, even when there is no case on point.187

A more forceful reform proposal would recognize that both certainty and predictability represent distinct, legitimate interests that can be realized and traded off against each other. For example, a court might favor plaintiffs’ interests in being made whole and deny qualified immunity whenever the court is certain that the plaintiffs have suffered a legal wrong. The court might then grant qualified immunity only if it is at least significantly unsure of the correct answer.188 That altered approach reflects a change in perspective as well as presumption: rather than finding qualified immunity unless the answer is obvious to all, the court would deny qualified immunity unless the court itself viewed the question as difficult. A court might therefore reject claims of qualified immunity not only when there is no case on point (as Jeffries and current doctrine would allow), but also when an official-defendant can cite some case law supportive of her position.

Finally, we again reach the issue of how to account for disagreements, particularly among courts. The Court has so far answered that question via prediction-oriented logic—and may have gone even further in accounting for dissensus in the qualified immunity context than when applying § 2254(d)(1).189 Unsurprisingly, the Court has accounted for circuit splits when ascertaining whether to grant qualified immunity.190 Thus, officials who act on an issue that is the subject of circuit disagreement tend to obtain qualified immunity, at least so long as no local precedent was on point.191 But the Court has also taken an added step and treated the existence of circuit disagreement on the issue of qualified immunity to be probative of whether to find qualified immunity.192 In other words, even when all judges have agreed on the merits, disagreement on whether that merits view was predictable has itself been treated as evidence of reasonable disagreement capable of supporting qualified immunity. That attention to clarity disagreements—that is, disagreements about clarity—goes beyond the attention to first-order disagreement on the merits typically seen in connection with § 2254(d)(1).193

Further, qualified immunity’s predictive aspect should differ from the analogous inquiry under § 2254(d)(1). In the federal habeas context, the predictability question is oriented around a state judicial actor, whereas qualified immunity applies to a much wider and generally less expert set of executive officials. That expertise differential suggests that the prediction standard should be more forgiving as well as more variegated in the qualified immunity context. For example, a federal court might plausibly conclude that police should have qualified immunity when there is any preexisting judicial precedent supportive of their view, since police would be hard pressed to second-guess or weigh competing judicial authorities. By contrast, a state court might reasonably be expected to appreciate that some judicial opinions, especially when isolated, are poorly reasoned or against the tide of more recent case law. Thus, there is reason to hold state court rulings to a higher standard of predictive sophistication and accuracy, even if that means choosing among competing legal authorities. Moreover, the qualified immunity inquiry might adapt to the specific executive official under review. Qualified immunity might be particularly forgiving when the relevant actor is, say, a police officer making a split-second decision, as opposed to an executive branch policymaker with access to an expert legal staff.194

Also differentiating qualified immunity from habeas, there are reasons to encourage executive officials to consider the views of many non-judicial actors. That point, too, flows from the fact that many officials lack legal expertise. The purposes of qualified immunity might call for encouraging those officials to seek and rely on relatively expert guidance.195 That point sometimes arises in current qualified immunity doctrine. For example, the Court has been more inclined to find qualified immunity when police act after consultation with relevant authorities, such as with judges issuing warrants.196 In a similar spirit, the Court has acknowledged the relevance of showing that police officers have acted in conformity with a local policy.197 And the Court has even smiled upon police who prepared a warrant application after seeking guidance from district attorneys—a factor that helped the police over and above the fact that they had obtained a judicial warrant.198 This line of consultative cases also supports the view that qualified immunity would almost automatically attach when any official acts in reliance on a memorandum prepared by the Office of Legal Counsel, an expert intra-executive adjudicatory body inside the Department of Justice.199

At the same time, there are also good reasons to prohibit judges from considering particular sources of information when making their decisions. A focus on predictability could thus support the creation of exclusionary principles that bar courts from entertaining evidence that might otherwise inform their views of legal correctness. This point cuts against a suggestion by Professors William Baude and Ryan Doerfler that judges in search of clarity might give weight to the views of like-minded epistemic peers, such as academic commentators who employ the same methodology as the judges themselves.200 That proposal focuses on the decisions of judges and attempts to improve their ability to form correct legal views. But if applied to qualified immunity, Baude and Doerfler’s approach to assessing legal clarity would significantly alter the incentives of executive branch decisionmakers. After all, executive officials act in the shadow of qualified immunity (and other doctrines). Even if judges are adept at finding commentators who are epistemic peers, police and other officials may not be. Inviting executive officials to make decisions based on approval from nongovernmental actors could invite greater error not by judges in their ultimate rulings, but by police and other executive officers in the field.

D. Constitutional Avoidance and Lenity

Clarity issues pervade statutory interpretation, as dozens of canons include clarity or ambiguity triggers. This Section explores these canons by focusing on what are arguably the two most salient examples: constitutional avoidance and lenity.201 Under the avoidance canon, ambiguous statutes are construed in favor of their constitutionality.202 And under the lenity canon, ambiguous statutes providing for criminal punishments are read narrowly.203 On reflection, these doctrines incorporate related but competing assumptions—and so might learn from one another, giving rise to several reform possibilities. The interaction between these canons also raises questions about the appeal of employing clarity doctrines at all in the context of statutory interpretation.204

Both lenity and constitutional avoidance operate on ambiguous statutes and so at first blush seem to share the same threshold requirement.205 But on further inspection, each canon can and does rest on a distinctive ambiguity trigger, such that a single statutory provision can be “ambiguous” for purposes of one doctrine, yet clear for the purposes of the other. Recent Court decisions portray lenity as being a rarely applied principle, pertinent only when there is a “grievous ambiguity,”206 whereas constitutional avoidance has found application in several cases where the relevant statutes could easily have been regarded as clear.207

In other words, the Court has embraced distinct clarity doctrines for different areas or aspects of statutory interpretation. And that basic result is not just unsurprising but entirely appropriate, so long as the relevant doctrines have relevantly different goals. But to work out whether the current doctrinal arrangement is in fact sensible, we need to identify the goals of those two doctrines, as well as how they might line up with the various forms of ambiguity we have discussed so far. One possibility is that both lenity and constitutional avoidance should focus on predictability—but, if so, the two contexts involve prediction by vastly different actors.

Start with lenity, in which the relevant actors are prospective criminal defendants. That group is heterogeneous, with widely varying abilities to predict legal outcomes. Some so-called “white collar” criminal statutes implicate business interests and so are more likely to apply to well-heeled defendants and corporations who have access to counsel and resources to mount a robust defense.208 Those counseled actors may be able to predict legal outcomes, including by staying apprised of relevant case law.209 By contrast, the great sweep of criminal law places burdens on legally unsophisticated individuals who have no realistic ability to apprehend relevant criminal statutes or obtain useful legal advice. Those actors—namely, regular people—steer clear of the criminal law based on a mix of common sense, conventional wisdom, and, if they are extraordinarily diligent, the law’s “plain meaning”—a term that could fruitfully be understood to encompass those aspects of statutory meaning that are readily accessible to a lay reader.210 The fact that most people enjoy only limited notice of the criminal law as an absolute matter, as well as reduced notice as a comparative matter, creates unfairness and so requires that courts adopt the relevant actors’ point of view.211 Attention to predictability further suggests that a heightened clarity threshold is particularly (and perhaps only) appropriate when interpreting criminal prohibitions outside the context of white-collar crime.212 Finally, a prediction-focused approach would place special weight on plain meaning, such as the conventional meaning of a statutory text. In sum, grounding lenity in the goal of promoting predictability sheds light on the canon’s scope, strength, and interaction with interpretive methods.

Constitutional avoidance can likewise be viewed as a function of predictability: when a legislature cannot reliably anticipate that its handiwork will suffer judicial invalidation, courts should strive to avoid that harsh result in favor of the purportedly softer sanction of reinterpretation.213 On that account of avoidance, the relevant actor is a legislative drafter. And, particularly in the context of federal legislation, drafters typically have access to abundant legal advice, including on matters of constitutional law. Thus, the expertise-based considerations that counseled in favor of frequently finding ambiguity under the doctrine of lenity are reversed for avoidance. So, to the extent that both lenity and constitutional avoidance rest on prediction, there is a strong case that the “grievous ambiguity” standard that currently applies to the lenity canon would be better suited in the context of avoidance, whereas the avoidance canon’s greater willingness to find legislative ambiguity would be better suited to lenity.214 The two doctrines, it seems, should swap clarity tests.

But perhaps one or both of these doctrines are actually best understood as being grounded in certainty, rather than predictability. This time, start with constitutional avoidance.

There are at least two ways to cast avoidance as focused on certainty, each with a parallel in other doctrines we have discussed. First, courts could engage in avoidance based on a theory of legislative intent and comparative expertise.215 Courts might plausibly think that democratically chosen and accountable legislatures know about their constitutional limitations and even have access to information about constitutionally relevant facts that courts generally lack.216 Given that assumption, a court might demand a high level of confidence before ruling that the legislature transgressed the Constitution. When only tentatively confident in its own assessment of legislative intent, the court might be persuaded to adopt a charitable view of the legislature’s work, at least if the court believes that the legislature has diligently exercised its constitutional expertise in a way that is relevant to the legal issue at hand. That approach somewhat parallels the comparative-expertise theory of Chevron, replacing the agency with the legislature.217

There is another way of viewing constitutional avoidance as resting on certainty: courts might require an especially high level of confidence about legislative intent before impinging on legally recognized interests or purposes. This approach—which I have referred to as the “New Holy Trinity”—would assume that statutory interpretation cases are about more than just getting legislative intent correct as often as possible.218 Other potential goals include: avoiding harm to people who have relied on legislation’s continued validity, preserving constitutional values like federalism, and safeguarding the Court’s institutional influence. If recognized as legally relevant, any of the foregoing goals could provide a reason not to follow apparent-but-uncertain legislative intent, even if fulfilling legislative intent is normally the preeminent objective of statutory interpretation. Imagine that a court is sure that a particular statutory reading would cause a harm and frustrate a legally recognized purpose but only modestly confident that that reading accords with legislative intent. The court’s low level of confidence regarding intent might lead it to prioritize its secondary goal of harm minimization. The doctrine of constitutional avoidance would then balance courts’ dueling commitments by requiring high confidence of legislative intent before adopting unconstitutional readings and thereby inflicting legally recognized harm.219

That certainty-based picture of constitutional avoidance points the way toward a certainty-based approach to lenity. Lenity cannot plausibly rest on an account of comparative expertise, since regular people have no discernible legal expertise that courts lack. But just as avoidance can be viewed as a requirement of caution before inflicting legally recognized harms, lenity can be viewed as a caution against inflicting potentially unwarranted harm via criminal punishment. Commentators have suggested that possibility,220 as have some of the Court’s rulings. 221 Note that this logic could find application with greater or lesser frequency, depending on how strongly courts value the interest in minimizing harm, relative to achieving legislative intent. If the Court is prepared to risk imposing criminal punishments more often than Congress intended, then it might implement its own best understanding of legislative intent, even when uncertain. Again, that is roughly the path the Court has taken.222

In sum, the Court’s recent enthusiasm for constitutional avoidance and aversion to lenity is probably best defended as the result of resting both those doctrines on certainty-based approaches. By contrast, greater attention to prediction-based considerations would likely lead to a reversal of how those two doctrines are implemented.

* * *

Investigation into constitutional avoidance and lenity casts light on broader patterns in the role of legal clarity in statutory interpretation. So many statutory interpretation doctrines are sensitive to clarity that the very practice of authoritatively reading statutes could be viewed as a complex process of navigating applicable clarity principles. Based on their differing clarity thresholds, the various doctrines arrange themselves into a decision tree or hierarchy. A court might begin with the “plain meaning” rule, which prescribes that clear text must control, regardless of other interpretative principles.223 Only when there is insufficient clarity to view a statute’s meaning as “plain” should courts turn to other doctrines governed by less stringent clarity thresholds. Consistent with that structured approach, the Court has essentially moved lenity to the end of the queue, considering it only after almost all other interpretive precepts.224

One might respond to the foregoing picture of statutory interpretation by suggesting that courts should simply think directly about the goals underlying lenity or avoidance, rather than inquiring into the presence of either clarity or ambiguity. After all, the critic might continue, these clarity doctrines’ ultimate goals do all the real analytical work, since they dictate the nature and severity of the relevant clarity thresholds. A court construing a criminal statute could simply ask itself whether its interpretation will unfairly surprise potential defendants, or a court encountering a possible constitutional defect in a statute could ask whether finding a violation would upset too many reliance interests—without ever asking as a discrete step whether the relevant laws are clear.

A possible rejoinder is that the terminology of clarity and ambiguity is simply how courts conventionally express their reasoning regarding underlying purposes. But that is not the whole story. By establishing case law on clarity’s practical meaning, courts routinize their decision-making. And in statutory interpretation as elsewhere, the law’s deep purposes often require distillation into standards or rules for their implementation to be workable, efficient, and fair.225 Clarity doctrines simply represent a particular type of implementation principle, one with distinctive advantages and risks. But inquiry into clarity, reasonable disagreement, or ambiguity is hardly the only analytical approach available to courts—and it may not be the best option, either. That leads to the next Part, which considers whether and when the clarity game is worth the candle.

III. Anxieties

So far, we have seen both that current doctrine is characterized by clarity pluralism and that at least a considerable degree of clarity pluralism is desirable, albeit not always in the form that current doctrine has chosen. But clarity pluralism is not inevitable, and clarity doctrines themselves are hardly an unalloyed good. Many legal mechanisms are worthy alternatives to inquiries into reasonable disagreement. For example, a court could eschew reliance on legal clarity in favor of applying a supermajority voting rule.226 And in at least some cases, it could make sense to abandon consideration of clarity. Without attempting to defend clarity doctrines against all comers, this Part discusses the most fundamental charges that could be leveled against having them at all.

A. Pluralism

We have now seen that the goals of any given clarity doctrine should and often do dictate the form of clarity sought under that doctrine. The result is clarity pluralism, or a diverse set of approaches to legal clarity in different doctrines. But even if each bespoke version of legal clarity makes sense in its unique doctrinal context, could the resulting clarity pluralism prove undesirably fragmented or confusing? If so, courts might foster the most obvious alternative: clarity uniformity.

There is some reason to think that courts do in fact resist clarity pluralism and foster uniformity instead. Clarity pluralism is often obscured by courts’ unelaborated use of the terms “clear,” “ambiguous,” and “reasonable” across different doctrines. While these courts may simply be overlooking doctrinal differences, their use of interchangeable terminology could be defended as a deliberate effort to link the relevant doctrines, allowing each to learn from others. For example, courts have expressly stated that the standard applicable in qualified immunity cases is identical to the “good faith” exception to the exclusionary rule, particularly in cases where police rely on warrants.227 As a result, any holdings under one line of cases instantly provide guidance in the other, hastening the law’s development.228 Those cases suggest that courts discern advantages in developing an “off the shelf” concept of legal clarity that can be applied in varied doctrinal contexts.

Yet the reasons for caring about legal clarity in the first place counsel against adopting completely uniform standards, and courts seem responsive to that pressure as well. As we have seen, clarity doctrines pursue varied ends, so a one-size-fits-all approach is unlikely to be attractive. The average of all the varied approaches to legal clarity would diverge too much and too often from the goals of each individual doctrine.229 And that problem obtains even across doctrines that adopt the same basic approach to legal clarity. We have seen, for example, that even if lenity and avoidance are both properly grounded in predictability, the relevant predictions would still be undertaken by radically different actors—namely, regular people and lawmakers.230

The most intuitive solution is to make uniform those clarity doctrines that share similar purposes. And it seems likely that the Court has done just that when self-consciously drawing comparisons between clarity doctrines. Again consider the Court’s decision to apply the same standard in cases when police rely on warrants, whether the issue arises in the context of the exclusionary rule or qualified immunity.231 Those two lines of case law exhibit obvious factual similarities, in that they both involve the same actors (police officers) engaged in the same recurring activity (use of warrants). But because the doctrines involve quite different remedies and claimants, there are plausible bases for distinguishing their clarity thresholds.232 For example, if the exclusionary rule were viewed as imposing greater social costs, then perhaps courts should apply a more demanding clarity threshold in exclusionary-rule cases, as opposed to qualified immunity cases. Yet the Court has concluded that both doctrines should aim to achieve roughly the same goal—namely, to protect even minimally competent police from the adverse consequences of their unlawful acts.233 Given that (eminently disputable) objective, it is unsurprising that the Court has generated similar prediction-oriented inquiries in both contexts.234 And by harmonizing the two doctrines, the Court has facilitated the law’s development while reinforcing a consistent message about the proper way for police to carry out their work.235

In sum, pluralism among clarity doctrines is not all good, nor is it inevitable. Courts experience a counter-pressure toward uniformity, with the result that pluralism is and should be checked, at least to some significant degree.

B. Malleability

We have seen that different judges often implement the same clarity doctrine differently, suggesting the possibility that legal clarity offers little more than an unprincipled “you know it when you see it” test. This problem is most acute when legal clarity rests on certainty, which in turn rests not on relatively objective indicia of predictability, but rather on often-disputed views of legal correctness. Again, background differences in methodology, such as adherence to textualism as opposed to purposivism, can yield divergences in how a certainty-based doctrine like Chevron is applied.236 So long as those sorts of background differences persist—and they will persist—there will also be differences in the implementation of certainty-based approaches to clarity. The most salient expositor of this worry is Justice Kavanaugh, whose views were canvassed earlier.237 Because courts use clarity doctrines without giving content to the relevant clarity thresholds, Justice Kavanaugh suggests the doctrines are, or at least appear, sufficiently malleable to achieve nearly any desired result.238

We have already seen one response to the malleability concern: because clarity doctrines are the products of law, courts can elaborate authoritative, rule-like clarity thresholds.239 And if courts’ views of legal correctness are too contested to foster outcome uniformity, a dollop of predictive clarity might do the trick.240 As we have already seen, Chevron could be revised so as to pay greater attention to widely held interpretive methodologies, as well as circuit splits and other objective evidence of whether judicial outcomes are foreseeable.241 But those responses could be rejected as unrealistic or for not sufficiently grappling with the problem. Because disputes about legal correctness bear on certainty-based clarity doctrines, convergence on the proper design of such a doctrine may prove elusive. More broadly, clarity doctrines obtain their content from underlying doctrinal purposes, so persistent disagreement as to those purposes can stymie progress at hashing out widely agreed-upon clarity doctrines. Put another way, clarity pluralism might abide not only across clarity doctrines, but also within them as jurists adopt differing approaches to the same doctrines, based on differing views of the legal goals at play.242 For example, the proper role of judicial deference to administrative agencies is a deep question, as evidenced by the now live debate on whether to throw Chevron overboard.243 That kind of deep disagreement is bound to influence how different judges apply Chevron.

Yet clarity doctrines sometimes create new opportunities for agreement on second-order decisional principles despite underlying disagreements on the first-order merits.244 Chevron is a potential example, as some formalists have approved it based on a theory of implicit delegation while functionalists appreciate that it accounts for agencies’ comparative expertise.245 Likewise, the plain meaning rule could be viewed as a practical compromise that reconciles competing views on the appeal of textualism: rather than generating opinions that inconsistently toggle between saying that text is either never or easily trumped by extrinsic evidence of legislative intent, the Court consistently posits that strong textual evidence, but only strong textual evidence, is definitive.246

Further, clarifying the nature of legal clarity would mitigate Justice Kavanaugh’s concern, even if the Supreme Court never adopted a clearer approach to legal clarity. Let us assume that greater understanding of how each judge understands and applies a given clarity doctrine does not yield a high level of outcome consistency across judges, which is Justice Kavanaugh’s express objective,247 so much as it reveals the reasons for persisting inconsistency. Even so, that improved understanding of judicial decision-making would mitigate Justice Kavanaugh’s worry that the informed public will view clarity doctrines as cynical or unprincipled. And once we understand how judges apply each doctrine, we can better assess whether the doctrine is on balance desirable. For example, even if Chevron were incapable of consistent application across judges due to ineradicable background disagreements about how to interpret statutes, Chevron might still have the effect of making agencies win more often than they otherwise would; and that outcome could very well be desirable. Further, a recent paper suggests that Chevron deference in fact tends to mute judges’ political biases.248 The prospect of attaining similar benefits could outweigh the necessarily attending costs and risks. Clarifying the underlying law of reasonable disagreement would enable courts and commentators alike to make those kinds of informed assessments.

C. Awkwardness

Clarity doctrines often require judges to sit in especially harsh judgment of one another, and the results can be awkward. Finding legal clarity could imply that a question was easy and that any interpreter to come out the other way was inept or subpar. This problem is especially acute under predictive approaches that set or enforce professional standards of competence. Judges who are averse to giving personal offense or undermining the courts’ legitimacy might accordingly veer away from finding clarity, even when clarity should be found. In time, clarity doctrines could become toothless, as judges fear insulting one another by declaring their views clearly wrong.

This dynamic may already have contributed to the trend toward increasingly high standards for legal clarity applied in qualified immunity and federal habeas corpus. Imagine that the justices are thinking about finding a violation of clearly established law under qualified immunity, but some lower courts have come out the other way on the merits. If the Court decreed that the law is clear, they would implicitly condemn their colleagues as unreasonable. Or imagine that a prominent executive official invoked qualified immunity: a court inclined to afford relief might have to declare her not just wrong but “plainly incompetent.”249 The Court’s qualified immunity cases exhibit just these anxieties, thereby dulling constitutional remedies.250 A similar problem arises under § 2254(d)(1): holding that a state court acted “unreasonably” sounds suspiciously like it calls for a declaration that state court judges acted incompetently. 251 Some federal judges might be averse to issuing that kind of statement, and that aversion may have contributed to the Court’s tendency to demand a breach of on-point precedent—that is, the most egregious form of judicial incompetence or insubordination—before invalidating a state-court conviction. Here too, the Court’s rhetoric, and even its reasoning, occasionally betrays just that worry.252

Once more, clarifying legal clarity can help. Establishing the reasons behind any given clarity doctrine would help judges understand and remember that decisions finding clear error advance important interests and so are worthwhile, despite their occasional awkwardness. More importantly, revised clarity doctrines can reduce or eliminate the harshness of finding clarity. Too often, the Court has equated “unreasonable” with “incompetent,” so that any judge or official who makes an “unreasonable” decision or violates “unambiguous” law must have acted shamefully.253 But incompetence is just one (rather extreme) standard that can give content to the idea of unreasonable legal views. When courts decree that a prior judicial ruling is “unreasonable” or, equivalently, that the relevant law is “clear,” they could at least sometimes explain that they are applying a complex analysis based on a variety of factors. And, if at least some attention to clarity-as-certainty enters into the picture, the relevant factors will include the deciding court’s own expert view of the legally correct answer, independent of any other decisionmaker’s competence.254 Once relevant actors understand how and why clarity findings arise, those findings would lose much of their stigmatic meaning and resulting distortive potential.

* * *

There is good reason to worry about the risk that some or all clarity doctrines are too pluralistic, malleable, and awkward. But clarifying the meaning of legal clarity can help mitigate those problems, in part by suggesting ways of adjusting current law and practice. Assessing each doctrine separately, courts can seek the optimal degree of clarity pluralism or uniformity, specify clarity tests and the potentially overlapping reasons for adopting them, and soften the harsh subtext that makes many clarity findings awkward. True, some legitimate anxieties would still remain, and perhaps courts should revise or abandon many clarity doctrines. But in taking that step, courts should keep in view the ways that clarity doctrines can and often do promote the law’s goals.

Conclusion

Clarity doctrines are doing well, as judged by the number and salience of decisions applying Chevron, qualified immunity, § 2254(d)(1), constitutional avoidance, lenity, the good faith exception to the exclusionary rule, and other legal principles that consider reasonable disagreement or ambiguity. But clarity doctrines are increasingly coming under scrutiny, including on the ground that they rest on conceptual sand. This Article has tried to answer a number of recurring questions about legal clarity and the related concepts of legal ambiguity and reasonable disagreement. To what extent do and should tests involving legal clarity operate consistently or varyingly across different doctrines? How can courts choose whether to flesh out legal clarity based on certainty or predictability? And how can courts decide how to set clarity thresholds and account for the existence of disagreement among legal interpreters? The answers to these questions may dictate not only the future shape of important clarity doctrines, but also whether some of these doctrines persist at all.

• 1The unadorned terms “clarity” and “ambiguity” here generally mean legal clarity and ambiguity, not their linguistic counterparts. As explained below, the law is often equated with the meaning of various legal texts, yielding a tendency to conflate indeterminacy about language with indeterminacy about law. Legal indeterminacy often results from linguistic indeterminacy, but they are not the same.
• 2See Chevron U.S.A. Inc v Natural Resources Defense Council, Inc, 467 US 837, 866 (1984).
• 3See, for example, United States v Bass, 404 US 336, 347 (1971).
• 4See, for example, Bond v United States, 572 US 844, 855 (2014).
• 5See, for example, Groh v Ramirez, 540 US 551, 563–65 (2004).
• 6See, for example, Terry Williams v Taylor, 529 US 362, 390–91 (2000).
• 7I follow the courts and lay practice in using the term “ambiguity” broadly to encompass various forms of indeterminacy or underdeterminacy, including vagueness. See Lawrence B. Solum, The Interpretation-Construction Distinction, 27 Const Commen 95, 97 (2010) (“In ordinary speech, the distinction between vagueness and ambiguity is not always observed . . . they both mark a general lack of what we might call ‘determinacy’ (or ‘clarity’ or ‘certainty’) of meaning.”). See also note 42 (discussing epistemic and metaphysical sources of uncertainty).
• 8Consider Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L J 511, 520 (noting that Chevron “suggests that the opposite of ‘ambiguity’ is not ‘resolvability’ but rather ‘clarity’”). In other words, to say a statute is ambiguous is to say it is unclear, not that it eludes an ultimate interpretive resolution.
• 9See Brett M. Kavanaugh, Book Review, Fixing Statutory Interpretation, 129 Harv L Rev 2118, 2144 (2016); Lawrence M. Solan, Pernicious Ambiguity in Contracts and Statutes, 79 Chi Kent L Rev 859, 859 (2004) (“The problem, perhaps ironically, is that the concept of ambiguity is itself perniciously ambiguous.”). See also Ward Farnsworth, Dustin F. Guzior, and Anup Malani, Ambiguity about Ambiguity: An Empirical Inquiry into Legal Interpretation, 2 J Legal Analysis 257, 276–77 (2010) (discussing the possibility of a general shift toward “external” assessments of ambiguity to avoid biased application of statutes). The Court has so far brushed aside worries about ambiguity’s ambiguity, though justices sometimes raise them. See text accompanying note 93. The idea that ambiguity is ambiguous is hardly new or limited to the law. See generally William Empson, Seven Types of Ambiguity (New Directions 1947) (a literary analysis).
• 10Throughout, I assume that courts will decide how to design or reform clarity doctrines. But legislators and other policymakers can of course do so as well, drawing on the framework here.
• 11As many commentators have put it: How clear is clear? See generally, for example, Note, “How Clear is Clear” in Chevron’s Step One?, 118 Harv L Rev 1687 (2005) (discussing Justice Scalia’s and others’ versions of this question).
• 12For an extraordinary example of how differently the same clarity doctrines can be applied, consider Judge Raymond Kethledge’s recent remark that, after almost ten years, his opinions have never found a statute ambiguous. See Raymond M. Kethledge, Ambiguities and Agency Cases: Reflections after (Almost) Ten Years on the Bench, 70 Vand L Rev En Banc 315, 320 (2017) (“In my own opinions as a judge, I have never yet had occasion to find a statute ambiguous.”).
• 13Professor Brian Slocum and Justice Brett Kavanaugh have made the main text’s points particularly forcefully. See Brian G. Slocum, The Importance of Being Ambiguous: Substantive Canons, Stare Decisis, and the Central Role of Ambiguity Determinations in the Administrative State, 69 Md L Rev 791, 806–07 (2010); text accompanying notes 144–48 (discussing Kavanaugh). See also Brian G. Slocum, Replacing the Flawed Chevron Standard, 60 Wm & Mary L Rev 195, 212–18 (2018).
• 14On the link between ambiguity and empirical conditions, consider the following passage:

  For example, whether criminal defendant X shot victim Y is an objective factual question—there is clearly a right answer in theory—but whether a court will be able to discern the correct answer depends on the availability of evidence. Similarly, ambiguity can cause uncertainty if there is insufficient evidence, but the uncertainty and disagreement will disappear if we have enough evidence.

  Tun-Jen Chiang and Lawrence B. Solum, The Interpretation-Construction Distinction in Patent Law, 123 Yale L J 530, 550 (2013).

• 15Commentators have made similar points in connection with particular doctrines. See Note, 118 Harv L Rev at 1688 (cited in note 11) (arguing that “the question ‘How clear is clear?’ should have a different answer depending upon the circumstances”). See also Jennifer E. Laurin, Trawling for Herring: Lessons in Doctrinal Borrowing and Convergence, 111 Colum L Rev 670, 702 (2011) (criticizing the “convergence” between qualified immunity and the exclusionary rule); Leah Litman, Remedial Convergence and Collapse, 106 Cal L Rev 1477, 1493–94 (2018) (criticizing convergence of the standards for clarity, among other things, adopted in doctrines like qualified immunity, Bivens, federal habeas corpus, and the exclusionary rule). As relevant here, Professor Leah Litman insightfully notes and critiques the “convergence” among clarity tests in several remediation doctrines, and the present analysis sheds added light on how that trend might be assessed. Still, significant clarity pluralism is evident even within the doctrines that Litman discusses. For example, under current case law, only reliance on clearly “on point” circuit precedent can (so far) trigger an exception to the exclusionary rule, whereas conflicted circuit precedent can support a finding of qualified immunity; and circuit precedent is never a basis for relief under 28 USC § 2254(d). Compare Davis v United States, 564 US 229, 241 (2011) (limiting the good-faith reliance on precedent in the exclusionary rule context to circumstances “when binding appellate precedent specifically authorizes a particular police practice”), with Wilson v Layne, 526 US 603, 618 (1999) (relying in part on “a split among the Federal Circuits” when granting qualified immunity) and Parker v Matthews, 567 US 37, 48–49 (2012) (“[C]ircuit precedent . . . cannot form the basis for habeas relief under AEDPA.”).
• 16From the leading treatment outlining this terminology:

  To say that a statute is ambiguous could be a claim that ordinary readers of English would disagree about its meaning, which we will call an external judgment. Or it could be a private conclusion that, regardless of what others might think, the reader is unsure how best to read the text—which we will call an internal judgment.

  Farnsworth, Guzior, and Malani, 2 J Legal Analysis at 258 (cited in note 9). Whereas Professor Ward Farnsworth, Dustin F. Guzior, and Professor Anup Malani cast external assessments of ambiguity in terms of “ordinary readers of English,” this Article emphasizes that prediction-oriented approaches can focus on any “external” interpreter, such as a state court judge or a police officer. Thus, this Article proceeds on the understanding that there are many plausible “external” perspectives that clarity doctrines might want to adopt.

• 17See text accompanying notes 208–12 (making this point in greater detail).
• 18See, for example, Caleb Nelson, What Is Textualism?, 91 Va L Rev 347, 396 (2005) (“[W]hen is one construction of a statute so superior to the alternatives that the administering agency has no option but to use it, and how close must the alternatives get in order to become ‘permissible’?”).
• 19See, for example, Cass R. Sunstein, Chevron as Law, 107 Georgetown L J *12–14, (forthcoming 2019) (discussing various degrees of clarity in percentile terms, such as 50–50, 90–10, and 60–40, without exploring the limitations of that shorthand). I confess that I, too, have used this shorthand. See Richard M. Re, Narrowing Supreme Court Precedent from Below, 104 Georgetown L J 921, 938 (2016).
• 20As we will see, Justice Kavanaugh has argued that confusion regarding percentile expressions of clarity supplies a reason to abandon clarity doctrines. See text accompanying notes 144–48 (addressing Justice Kavanaugh’s frustrations with the Chevron doctrine).
• 21Consider Frank H. Easterbrook, The Role of Original Intent in Statutory Construction, 11 Harv J L & Pub Pol 59, 62 (1988) (“There is no metric for clarity.”).
• 22See Kavanaugh, Book Review, 129 Harv L Rev at 2144 (cited in note 9).
• 23See note 226 and accompanying text.
• 24The last two doctrines are discussed en passant, particularly within Parts II.D and III.A.
• 25See note 7 (discussing use of the term “ambiguity”).
• 26See, for example, In re Rogers, 513 F3d 212, 226 (5th Cir 2008) (“For the language to be considered ambiguous, however, it must be susceptible to more than one reasonable interpretation or more than one accepted meaning.”); Resource Bankshares v St. Paul Mercury Insurance Co, 407 F3d 631, 636 (4th Cir 2005) (“A policy provision is ambiguous when, in context, it is capable of more than one reasonable meaning.”); O’Neil v Retirement Plan for Salaried Employees of RKO General, Inc, 37 F3d 55, 59 (2d Cir 1994) (“Language is ambiguous when it is capable of more than one meaning when viewed objectively by a reasonably intelligent person who has examined the context of the entire integrated agreement.”) (quotation marks and citation omitted).
• 27Clarity is often portrayed as a high degree of correctness, such as in the expression “It’s not just correct—it’s clear!” But clarity can incorporate or consist of traits that are independent of correctness. For example, clarity might exist when an interpretation is not just correct, but widely agreed upon. So defined, correctness and clarity would not be the same type of trait.
• 28Clarity could be taken to require perfect correctness, ease, confidence, and agreement, in which case finding clarity based on any subset of those variables would amount to settling for imperfect clarity. For a statement on what could be regarded as perfect clarity, see Brian Bix, Law, Language, and Legal Determinacy 63 (Clarendon 1993) (“[L]egal questions that seem so obvious that there is no question as to what their proper resolution should be” are sometimes labeled “clear cases.”). Because perfectly clear issues are unlikely to reach courts, see Frederick Schauer, Easy Cases, 58 S Cal L Rev 399, 411–14 (1985), clarity doctrines can have significant effect only by dealing in imperfection.
• 29In other words, clarity doctrines can treat certain actual or hypothetical interpreters as a baseline or reference point. Consider René Descartes, Principles of Philosophy 1–45 in Descartes: Selected Philosophical Writings 160–74 (Cambridge 1984) (John Cottingham, Robert Stoothoff, and Dugald Murdoch, trans) (asserting that perceptions are “clear” when they are “present and accessible to the attentive mind”) (emphasis added).
• 30For example, a proposition may be clear for purposes of qualified immunity, given an officer’s perspective on circuit case law, even if the opposite view is correct and ultimately prevails in the Supreme Court. On that possibility, see note 169. See also text accompanying note 39 (providing an example of clear but incorrect prediction).
• 31See, for example, Mark Greenberg, Legislation as Communication? Legal Interpretation and the Study of Linguistic Communication, in Andrei Marmor and Scott Soames, eds, Philosophical Foundations of Language in the Law 217, 226 (Oxford 2011), citing Smith v United States, 508 US 223 (1993). See also Mark Greenberg, The Standard Picture and Its Discontents, in 1 Oxford Stud in Phil of Law 39, 49–50 (2011) (arguing that the widely held “standard picture” is deficient in that it conflates linguistic and legal meaning); Mark Greenberg, Response, What Makes a Method of Legal Interpretation Correct?: Legal Standards vs. Fundamental Determinants, 130 Harv L Rev F 105, 107–10 (2017) (emphasizing “the crucial distinction between linguistic meaning and the content of the law”).
• 32See note 26. See also Solan, 79 Chi Kent L Rev at 860 (cited in note 9) (noting that “linguists and philosophers” define ambiguity as “an expression [that] can be understood in more than one distinct sense,” but that “[l]egal writers, and judges in particular, use the word ‘ambiguity’ to refer to all kinds of indeterminacy, whatever their source”).
• 33See generally William Baude and Stephen E. Sachs, The Law of Interpretation, 130 Harv L Rev 1079 (2017) (drawing in part on Professor Greenberg’s work in discussing legal canons and other closure rules). The present Article explores the “law of clarity.”
• 34See, for example, Eric A. Posner and Adrian Vermeule, The Votes of Other Judges, 105 Georgetown L J 159, 178 (2016) (noting a variety of potential numerical probabilities and positing, “A judge with a high confidence level believes that the statute is clear; a judge with a confidence level in the neighborhood of 0.5 believes that the statute is ambiguous”); note 19 (quoting Sunstein); text accompanying notes 144–45 (quoting Justice Kavanaugh).
• 35See note 34.
• 36We could additionally imagine an impersonal perspective, from which legal issues are deemed clear or unclear based on their inherent properties without regard to how any particular interpreter would view the matter. But because that kind of “view from nowhere” is elusive and obscure, clarity doctrines have good reason to adopt a more particularized perspective, whether of an actual or a constructed interpreter.
• 37Certainty and predictability could be viewed as consequences of clarity, and their presence in any given context would then be evidence that clarity is present too. For example, if clarity is defined as the ease of accessing a legal conclusion, then clear answers will tend to yield both certainty and predictability. On that view, the key question is which type (or types) of evidence to privilege.
• 38Interpretive ease and certainty are correlated, but they can come apart. For an approach that focuses on ease or efficiency as opposed to confidence, see text accompanying note 56.
• 39Consider Thomas C. Schelling, The Strategy of Conflict 54–56 (Harvard 1960) (using a similar example to illustrate tacit coordination). Similarly, a statute’s text might immediately suggest a meaning that many interpreters would converge on but that better resourced interpreters can show to be debatable or even incorrect. Lagos v United States, 138 S Ct 1684 (2018), might offer an example: after a string of circuits embraced the same reading of a statute, the DC Circuit elaborated a contrary reading that the Supreme Court unanimously adopted.
• 40For example, a clarity doctrine could attempt to instill a widely applicable rule, such as, “The presence of a circuit split always generates unclarity for purposes of this doctrine.” Or a clarity doctrine could outline a number of factors as a standard for ascertaining clarity in each case where the doctrine comes to bear.
• 41In a new paper, Professor Ryan Doerfler explores clarity doctrines related to statutory interpretation (as opposed to case law) and helpfully distinguishes between “evidence-management doctrines” that structure inquiry into statutory meaning and “uncertainty-management doctrines” that help courts “play it safe.” Ryan D. Doerfler, Going “Clear” *8–16, 27 (University of Pennsylvania Law School Research Paper No 19-10, Jan 2019), archived at http://perma.cc/W3QJ-ARJ4. Doerfler’s categories represent two available subtypes of what I describe below as certainty-based clarity, in that he assumes the perspective of the deciding court and explores either how certainty arises or what to do once the court has arrived at whatever form of certainty is available.
• 42This point goes to the difference between uncertainty that stems from epistemic limitations as opposed to metaphysical indeterminacy: Is the right answer simply inaccessible, or is there no answer at all? In a sense, uncertainty that stems from indeterminacy could be regarded as a kind of certainty—namely, a certainty that no answer exists. Consider Ken Kress, A Preface to Epistemological Indeterminacy, 85 Nw U L Rev 134, 138 (1990) (“Metaphysical indeterminacy speaks to whether there is law; epistemic indeterminacy, to whether the law can be known.”); Kent Greenawalt, Discretion and Judicial Decision: The Elusive Quest for the Fetters That Bind Judges, 75 Colum L Rev 359, 378 (1975) (finding that “when more than one result will widely be regarded as a satisfactory fulfillment of his judicial responsibilities then . . . as far as the law is concerned, he has discretion to decide between them”). These two kinds of certainty-based clarity operate in tandem. At the metaphysical level, legal clarity exists when there is a determinate legal answer to be discovered. And, at an epistemic level, legal clarity exists when there is appropriately high confidence as to what the one legally available answer is.
• 43See Farnsworth, Guzior, and Malani, 2 J Legal Analysis at 258 (cited in note 9).
• 44Professor Farnsworth, Guzior, and Professor Malani take valuable first steps in this direction by suggesting that, given their relative purposes, Chevron and lenity might respectively be more concerned with internal and external assessments of ambiguity. Id at 281–83. However, the authors contend that the “advantages of external inquiries into ambiguity cannot and should not be formalized into a rule for legal use.” Id at 290. These possibilities are explored in more detail in Part II below.
• 45For example, Brand X scenarios—wherein an agency view displaces an earlier judicial elaboration of statutory meaning—can be viewed as new information for courts applying Chevron deference, on a comparative expertise theory. See National Cable & Telecommunications Association v Brand X Internet Services, 545 US 967, 982 (2005). See also Part II.B.
• 46The Court sometimes defers decision on difficult issues in part to gather additional information before acting. See, for example, City of Ontario, California v Quon, 560 US 746, 759 (2010) (“The judiciary risks error by elaborating too fully on the Fourth Amendment implications of emerging technology before its role in society has become clear. . . . Prudence counsels caution . . . .”). Clarity doctrines can facilitate that deferral.
• 47Again, legal indeterminacy could represent certainty that no answer exists. See note 42.
• 48See Kavanaugh, Book Review, 129 Harv L Rev at 2142 (cited in note 9). See also Gary Lawson, Legal Indeterminacy: Its Cause and Cure, 19 Harv J L & Pub Pol 411, 411 (1995) (“[O]ne needs to know how much uncertainty is enough to create indeterminacy.”). Professor Gary Lawson’s point about indeterminacy extends to unclarity. Lawson also makes the important point that not just standards of proof but also burdens of proof can help dictate the existence of indeterminacy. Id at 423.
• 49Consider Daryl J. Levinson, Rights Essentialism and Remedial Equilibration, 99 Colum L Rev 857, 915 (1999) (suggesting that a stricter and clearer rule for reasonableness in Fourth Amendment claims would reduce police officers’ ability to rely on qualified immunity).
• 50As we will see, doctrine has generally followed that second path. See Part II.C.
• 51See, for example, text accompanying notes 178 and 187.
• 52By contrast, the Court has suggested that there is no clearly established law under the Anti-Terrorism and Effective Death Penalty Act of 1996 (AEDPA) when the claimant identifies “the logical next step from” existing case law, so long as that step has “not yet” been “taken” and “there are reasonable arguments on both sides.” White v Woodall, 572 US 415, 427 (2014).
• 53Professors Eric Posner and Adrian Vermeule have proposed largely the picture outlined in the main text, and they further contend that disagreements among judges should strongly counsel in favor of finding unclarity. Posner and Vermeule, 105 Georgetown L J at 163 (cited in note 34). See also Alex Stein, Law and the Epistemology of Disagreements, 96 Wash U L Rev 51, 81–89 (2018).
• 54See Richard M. Re, The New Holy Trinity, 18 Green Bag 2d 407, 415 (2015). See also Ryan D. Doerfler, High-Stakes Interpretation, 116 Mich L Rev 523, 564 (2018).
• 55Consider John Rawls, A Theory of Justice 150–57 (Harvard 1971) (discussing a maximin approach to risk distribution as to the basic structure of society). Rawls did not argue for a maximin approach to legal interpretation. See also William N. Eskridge Jr, Public Values in Statutory Interpretation, 137 U Pa L Rev 1007, 1032 (1989) (discussing a canon by which statutes are construed in favor of discrete and insular minorities).
• 56Qualified immunity currently calls for a kind of cost-benefit analysis when courts decide whether to reach the merits in cases in which the law is unclear. See Pearson v Callahan, 555 US 223, 236–42 (2009). We could imagine altering the doctrine such that, if (and only if) the cost-benefit analysis justifies a merits holding, then the court should declare the law clear.
• 57But see Doerfler, 116 Mich L Rev at 542 (cited in note 54) (“This Article’s working hypothesis is that to say that it is ‘clear’ (or ‘plain’ or ‘unambiguous’) that something is the case is, roughly speaking, to claim that one is in an epistemic position to ‘know’ it.”).
• 58This approach is reminiscent of Skidmore deference. Skidmore v Swift & Co, 323 US 134, 140 (1944) (arguing that the “weight” of an agency view “will depend upon the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, if lacking power to control”). Though not styled as a clarity doctrine, Skidmore could be re-described as one: whether a court applying Skidmore views an issue as “unclear” and so allows an agency’s view to resolve it might depend in part on whether the agency is sufficiently credible, persuasive, and so forth.
• 59See, for example, Thomas Kelly, Peer Disagreement and Higher-Order Evidence, in Richard Feldman and Ted A. Warfield, eds, Disagreement 111, 164 (Oxford 2010) (“[I]n some cases, it might very well be rational for you to conclude that your friend is not your peer after all, where your only basis for so concluding is the lack of judgment that he displays in subsequent cases in which the two of you disagree.”).
• 60By analogy, someone might seem like an epistemic peer until he inaccurately sums two and two to obtain five––and then persists in defending that conclusion. See generally William Baude and Ryan D. Doerfler, Arguing with Friends, 117 Mich L Rev 319 (2018) (arguing that judges should adjust their own assessments of legal questions by taking into account the views of methodologically like-minded interpreters who are epistemic peers, not other interpreters).
• 61As Justice Stephen Breyer has put in the context of Chevron deference, “It is difficult, after having examined a legal question in depth with the object of deciding it correctly, to believe both that the agency’s interpretation is legally wrong, and that its interpretation is reasonable.” Stephen Breyer, Judicial Review of Questions of Law and Policy, 38 Admin L Rev 363, 379 (1986). However, Breyer’s important observation is not equally true of all clarity doctrines.
• 62See, for example, Daniel Kahneman and Amos Tversky, On the Reality of Cognitive Illusions, 103 Psychological Rev 582, 587–89 (1996).
• 63See Farnsworth, Guzior, and Malani, 2 J Legal Analysis at 271–72 (cited in note 9) (“[R]espondents with strong policy preferences are, on average, 30 percent more likely to say a statute is ambiguous or probably ambiguous when asked for an external judgment than when asked for an internal one.”).
• 64See Parts II.B (Chevron) and III.B (Malleability).
• 65Consider Richard A. Posner, Legal Formalism, Legal Realism, and the Interpretation of Statutes and the Constitution, 37 Case W Res L Rev 179, 187 (1986) (“A text is clear if all or most persons, having the linguistic and cultural competence assumed by the authors of the text, would agree on its meaning.”).
• 66Part II discusses clarity doctrines that involve a variety of outside actors, such as courts (AEDPA), executive officials (qualified immunity), agencies (Chevron), legislators (avoidance), and private parties (lenity).
• 67See, for example, United States v Lanier, 520 US 259, 266 (1997) (“[T]he canon of strict construction of criminal statutes, or rule of lenity, ensures fair warning by so resolving ambiguity in a criminal statute as to apply it only to conduct clearly covered.”); Hope v Pelzer, 536 US 730, 740 (2002) (explaining that, under Lanier, “the ‘fair warning’ requirement is identical under [18 USC] § 242 and the qualified immunity standard”).
• 68Without various clarity doctrines pertaining to remedies, the just-mentioned concerns about unfairness and disruptiveness might discourage courts from reaching unpredicted results at all, even when those results would be more accurate or socially beneficial. See John C. Jeffries Jr, The Right-Remedy Gap in Constitutional Law, 109 Yale L J 87, 113 (1999) (“The cost of innovation is held down if new rules can be adopted without full compensation for the discarded past practice.”).
• 69See note 16. For a description of the objectivity and debiasing benefits associated with external assessments of ambiguity, see Farnsworth, Guzior, and Malani, 2 J Legal Analysis at 272 (cited in note 9) (“When respondents are asked whether ordinary readers of English would be likely to agree on the best reading of the statute in that case, their judgments are unaffected by their policy preferences.”). But even if the particular form of externality that the authors tested is indeed uniquely unbiased, it still might not capture considerations that are doctrinally relevant. For example, imagining the views of “ordinary readers of English” might prevent consideration of canons and other closure rules accessible to relevant interpreters, such as courts.
• 70Or, to borrow Professor Scott Dodson’s apt question: “Clear to whom?” See Scott Dodson, The Complexity of Jurisdictional Clarity, 97 Va L Rev 1, 4 (2011).
• 71See Part II.D (lenity).
• 72The doctrine of qualified immunity has increasingly shifted toward this type of justification as it has become more concerned with unpredictability. See Part II.C. See also, for example, City and County of San Francisco v Sheehan, 135 S Ct 1765, 1774 n 3 (2015) (emphasizing “the importance of qualified immunity to society as a whole” because of its asserted liberating effects on officer behavior) (quotation marks omitted).
• 73For example, we will see that § 2254(d)(1) “restricts the source of clearly established law to this Court’s jurisprudence.” Terry Williams, 529 US at 412. See text accompanying note 88.
• 74See Parts II.A (AEDPA) and II.C (qualified immunity).
• 75See notes 223, 246, and accompanying text.
• 76See, for example, Hernández v Mesa, 137 S Ct 2003, 2007 (2017) (providing an example from qualified immunity case law).
• 77Qualified immunity doctrine has so far resisted this kind of evaluation of officer decision-making process, though it has been criticized on that ground. See Brandon Garrett and Seth Stoughton, A Tactical Fourth Amendment, 103 Va L Rev 211, 217–18, 263 (2017) (“Fourth Amendment decisions do not take any particular notice of the tactics that are designed to maximize the quality of officer decision making.”); Daniel Epps, SCOTUS Symposium: Thoughts on County of LA v. Mendez (PrawfsBlawg, May 20, 2017), archived at http://perma.cc/JD7F-8MQJ. A more sophisticated but still prediction-oriented approach might withhold qualified immunity when police rush to act even though caution is called for.
• 78This possibility comes up below in connection with lenity. See Part II.D.
• 79Offering a dramatic illustration of the dynamic suggested in the main text, Professor Mila Sohoni has argued that the New Deal Court lowered the clarity thresholds for a variety of apparently predictive doctrines, including vagueness and lenity, partly in order to make room for greater legislative authority and the creation of the administrative state. See generally Mila Sohoni, Notice and the New Deal, 62 Duke L J 1169 (2013).
• 80Consider Stanton v Sims, 571 US 3, 9–10 (2013) (finding it “especially troubling” that an officer would be deemed “plainly incompetent—and subject to personal liability for damages—based on actions that were lawful according to courts in the jurisdiction where he acted”).
• 81See Baude and Doerfler, 117 Mich L Rev at 323 (cited in note 60) (arguing that courts in search of legal clarity should focus attention on “epistemic peers,” defined in part as “individuals who are equally likely to get things right (or wrong) with respect to a given issue”). See also Posner and Vermeule, 105 Georgetown L J at 171–72 (cited in note 34) (in effect arguing that judges are usually one another’s epistemic peers).
• 82See Part III.C (Awkwardness).
• 83For example, Professors Posner and Vermeule characterize the clarity doctrines associated with federal habeas corpus, Chevron, qualified immunity, and many others, as “complex” doctrines that should at least presumptively be governed by similar principles. See Posner and Vermeule, 105 Georgetown L J at 162–70 (cited in note 34).
• 84Pub L No 104-132, 110 Stat 1214 (1996), codified as amended in various sections of Title 28.
• 8528 USC § 2254(d)(1). Section 2254(d)(2) allows for relief when a state adjudication “resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.”
• 86529 US 362 (2000).
• 87Id at 405–08.
• 88Id at 412 (emphasis added).
• 89Harrington v Richter, 562 US 86, 103 (2011) (emphasis added).
• 90See Terry Williams, 529 US at 377–79 (opinion of Stevens). Justice Stevens announced the judgment of the Court and part of his opinion—on relatively inconsequential matters—was the opinion for the Court. When it came to the key legal standard discussed in the main text, Justice Stevens in effect dissented.
• 91Id at 389.
• 92See also id (asserting that “state-court judgments must be upheld unless, after the closest examination of the state-court judgment, a federal court is firmly convinced that a federal constitutional right has been violated”) (emphasis added).
• 93Id at 410.
• 94Terry Williams, 529 US at 410.
• 95In fairness to Justice O’Connor, some statements in Justice Stevens’s opinion could be read as entirely rejecting deference to state court judgments. See Terry Williams, 529 US at 387 (opinion of Stevens) (“Whatever ‘deference’ Congress had in mind with respect to both phrases [in § 2254(d)(1)], it surely is not a requirement that federal courts actually defer to a state-court application of the federal law that is, in the independent judgment of the federal court, in error.”). As we have seen, however, Justice Stevens’s opinion ultimately required a “firm conviction” of error, not just a belief. See notes 91–92 and accompanying text.
• 96The majority also argued from various traditional sources of legal interpretation, but none actually pointed toward any particular view of clarity. For example, Justice O’Connor insisted on making meaningful the statute’s separate “contrary to” and “unreasonable application of” prongs. Id at 412–13. And Justice O’Connor emphasized that Congress wanted to override a prior ruling on “whether . . . a federal habeas court should ask whether the state-court decision was correct or simply whether it was reasonable.” Id at 411. The Court also cited legislative history that, as Stevens noted, does little more than restate the statutory language.
• 97See Virginia v LeBlanc, 137 S Ct 1726, 1729 (2017) (per curiam); Wainwright v Sykes, 433 US 72, 90 (1977). Related to finality, the Court has long worried that expansive federal habeas review would tend to make state-court trials a mere “tryout on the road” for later federal habeas adjudication, thereby fostering adjudicative redundancy and delay while diminishing local control over criminal justice. Wainwright, 433 US at 90.
• 9828 USC § 2254(d)(1). By contrast, circuit precedent can play a more central role when adjudicating qualified immunity claims. See note 169.
• 99Harrington, 562 US at 102–03 (quotation marks omitted).
• 100See, for example, Terry Williams, 529 US at 378–79 (opinion of Stevens) (“At the core of this power is the federal courts’ independent responsibility . . . to interpret federal law.”).
• 101See note 95 and accompanying text (acknowledging that Justice Stevens sometimes seemed to repudiate any deference at all).
• 102We will discuss an analogous reform for qualified immunity. See text accompanying note 188.
• 103541 US 652 (2004).
• 104Id at 665–66.
• 105564 US 261 (2011).
• 106Id at 265.
• 107Alvarado, 541 US at 666. The court below had held that it should “extend a clearly established legal principle [of the relevance of juvenile status] to a new context.” Id (quotation marks omitted).
• 108The Court concluded that its holding flowed from precedent, “commonsense,” and “clear” practical implications. J.D.B., 564 US at 271–72.
• 109Alvarado, 541 US at 675 (Breyer dissenting) (“As I have said, the law in this case is clear.”).
• 110See Brecht v Abrahamson, 507 US 619, 636 (1993) (noting that “state courts often occupy a superior vantage point from which to evaluate the effect of trial error”).
• 111Attention to the magnitude of the relevant constitutional harm could also support adjusting the AEDPA clarity standard to account for the severity of the punishment at issue. But see White v Wheeler, 136 S Ct 456, 462 (2015) (emphasizing that “the provisions of AEDPA apply with full force even when reviewing a conviction and sentence imposing the death penalty”).
• 112Terry Williams, 529 US at 386 (opinion of Stevens) (emphasis added). Stevens’s remark could also suggest an efficiency-oriented approach to legal clarity: only a sufficiently serious error would justify the effort of revisiting a final criminal conviction. See text accompanying note 56.
• 113See Woods v Etherton, 136 S Ct 1149, 1151 (2016) (“The state court decision must be ‘so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement.’”) (citation omitted).
• 114White v Woodall, 572 US 415, 419 (2014) (alteration and citation omitted).
• 115Consider Pearson v Callahan, 555 US 223, 239 (2009) (outlining a similar inquiry in the qualified immunity context).
• 116Terry Williams, 529 US at 409 (citation omitted).
• 117Id at 410.
• 118Id at 378 (opinion of Stevens).
• 119Carey v Musladin, 549 US 70, 76 (2006) (“Reflecting the lack of guidance from this Court, lower courts have diverged widely in their treatment of defendants’ spectator-conduct claims.”). See also White v Woodall, 572 US at 422 n 3 (“The Courts of Appeals have recognized that Mitchell left this unresolved; their diverging approaches to the question illustrate the possibility of fairminded disagreement.”); McWilliams v Dunn, 137 S Ct 1790, 1804 (2017) (Alito dissenting). See also Ruth A. Moyer, Disagreement about Disagreement: The Effect of a Circuit Split or “Other Circuit” Authority on the Availability of Federal Habeas Relief for State Convicts, 82 U Cin L Rev 831, 857 (2014); Noam Biale, Beyond a Reasonable Disagreement: Judging Habeas Corpus, 83 U Cin L Rev 1337, 1371 (2015).
• 120But see Aaron-Andrew P. Bruhl, Following Lower-Court Precedent, 81 U Chi L Rev 851, 890 (2014) (concluding there are circumstances in which the Court is more likely to defer to lower-court precedent).
• 121Epistemic and practical differences between decision-making by the justices (as opposed to lower-court judges) may counsel in favor of implementing clarity doctrines differently depending on which court is applying them. See also note 158 (discussing Chevron and hierarchical differences within the judiciary).
• 122See Lopez v Smith, 574 US 1, 4 (2014) (per curiam).
• 123But see Marshall v Rodgers, 569 US 58, 64 (2013) (insisting that a circuit court “may not canvass circuit decisions to determine whether a particular rule of law is so widely accepted among the federal circuits that it would, if presented to this Court, be accepted as correct”); Parker v Matthews, 567 US 37, 49 (2012) (denying that “the Sixth Circuit’s reliance on its own precedents [can] be defended in this case on the ground that they merely reflect what has been ‘clearly established’ by our cases”).
• 124467 US 837 (1984).
• 125Id at 842–44.
• 126Id at 842–43.
• 127Id at 844.
• 128Profesor Farnsworth, Guzior, and Professor Malani make the related point that Chevron ambiguity might rest on intentionalist considerations rather than textual ones: “In this case the intent of Congress, rather than the public meaning of the text, probably is the more important benchmark if there is a conflict between them, because giving effect to the intent of Congress is the most widely accepted rationale for the Chevron doctrine.” Farnsworth, Guzior, and Malani, 2 J Legal Analysis at 281 (cited in note 9).
• 129Chevron, 467 US at 843 n 9.
• 130See Frederick Liu, Chevron as a Doctrine of Hard Cases, 66 Admin L Rev 285, 291 (2014) (suggesting that “a statute is clear when the law provides an answer before running out”). Liu essentially views clarity as determinacy.
• 131If the court upholds an agency interpretation as unambiguously correct, by contrast, then no other view is permissible, even if the agency later changes its position. See National Cable & Telecommunications Association v Brand X Internet Services, 545 US 967, 982 (2005). See also United States v Home Concrete & Supply, LLC, 566 US 478, 488 (2012) (plurality). For discussion of how Brand X comports with a certainty-based approach, see note 45.
• 132See, for example, Wilson v Layne, 526 US 603, 613–15 (1999).
• 133See Chevron, 467 US at 843–44 (discussing the “legislative delegation to an agency”); Cass R. Sunstein, Chevron Step Zero, 92 Va L Rev 187, 192 (2006). Under an implied delegation theory, Chevron arguably implements a kind of knowledge—namely, knowledge that Congress authorized the agency to resolve the relevant legal issue. See note 42 (discussing certainty that there is legal indeterminacy).
• 134See Chevron, 467 US at 865 (noting the “great expertise” of agencies). See also Scalia, 1989 Duke L J at 514–15 (cited in note 8) (critically discussing the view, present in both “old and new” cases, that agencies “are more likely than the courts to reach the correct result” in part because of “their intense familiarity with the history and purposes of the legislation at issue” and further noting that “[p]olicy evaluation is . . . part of the traditional judicial tool-kit that is used in applying the first step of Chevron,” including when finding absurdity or discovering legislative understandings).
• 135See Exxon Mobil Corp v Allapattah Services, Inc, 545 US 546, 572 (2005) (Stevens dissenting) (noting that Justice Ruth Bader Ginsburg’s separate dissent “has demonstrated that ‘ambiguity’ is a term that may have different meanings for different judges” and that “ambiguity is apparently in the eye of the beholder”). This point underwrites Justice Kavanaugh’s critique of Chevron.
• 136See Robert A. Katzmann, Response to Judge Kavanaugh’s Review of Judging Statutes, 129 Harv L Rev F 388, 398 (2016) (“Wouldn’t resorting to reliable legislative history act as a restraint on a judge’s substituting preferences for that of the legislature . . . ?”).
• 137But see Kent Barnett, Christina L. Boyd, and Christopher J. Walker, Administrative Law’s Political Dynamics, 71 Vand L Rev 1463, 1468–69 (2018) (arguing based on an empirical study that Chevron deference generally mutes the effects of judicial partisanship and fosters uniform outcomes across judges).
• "</a138">See Scalia, 1989 Duke L J at 521 (cited in note 8) (“One who finds more often (as I do) that the meaning of a statute is apparent from its text and from its relationship with other laws, thereby finds less often that the triggering requirement for Chevron deference exists.”). See also Kavanaugh, Book Review, 129 Harv L Rev at 2129 (cited in note 9) (“[T]extualists tend to find language to be clear rather than ambiguous more readily than purposivists do.”), citing Scalia, 1989 Duke L J at 521 (cited in note 8) and Elena Kagan, The Scalia Lecture: A Dialogue with Justice Kagan on the Reading of Statutes at 56:54 (Nov 17, 2015), online at http://www.youtube.com/watch?v=dpEtszFT0Tg (visited Mar 26, 2019) (Perma archive unavailable) (noting differences between herself and Justice Scalia as to “the quickness with which we find ambiguity”); Kavanaugh, Book Review, 129 Harv L Rev at 2152 (cited in note 9) (“Chevron is not determinate because it depends on the threshold clarity versus ambiguity determination. As Justice Scalia pointed out, that determination is the chink in Chevron’s armor—the ambiguity that prevents it from being an absolutely clear guide to future judicial decisions.”) (quotation marks omitted).
• 139What the main text describes could be viewed as a relatively subtle version of the so-called “major questions” doctrine, which is typically portrayed as an exception to Chevron rather than a means of calibrating it. However, the Court sometimes describes the doctrine in calibration-like terms. See, for example, FDA v Brown & Williamson Tobacco Corp, 529 US 120, 159 (2000) (“In extraordinary cases, however, there may be reason to hesitate before concluding that Congress has intended such an implicit delegation.”).
• 140We can further imagine varying Chevron’s clarity threshold on an agency-by-agency basis, rather than in a uniform way across agencies, depending on each agency’s track record at any given moment. Such an adjustment would move Chevron toward one conventional understanding of Skidmore. Consider Peter L. Strauss, “Deference” Is Too Confusing—Let’s Call Them “Chevron Space” and “Skidmore Weight”, 112 Colum L Rev 1143, 1153–56 (2012); note 58 (discussing the possibility of recasting Skidmore as a kind of clarity doctrine).
• 141Chevron, 467 US at 842–43.
• 142Id at 843 n 9.
• 143Slocum, 69 Md L Rev at 829 (cited in note 13).
• 144See Kavanaugh, Book Review, 129 Harv L Rev at 2150 (cited in note 9).
• 145Id at 2137.
• 146Id at 2140, 2142. See also United States v Yermian, 468 US 63, 77–78 (1984) (Rehnquist dissenting) (complaining that the Court used “the magic wand of ipse dixit” in finding ambiguity).
• 147Four justices, including Justice Kavanaugh himself, recently drew on this claim in criticizing Auer v Robbins, 519 US 452 (1997), which calls for judicial deference to administrative agencies’ reasonable interpretations of their own regulations. See Kisor v Wilkie, 139 S Ct 2400, 2425 (2019) (Gorsuch concurring in the judgment).
• 148Kavanaugh, Book Review, 129 Harv L Rev at 2137 (cited in note 9).
• 149See Part II.A.
• 150See Part I.A.
• 151See Chevron, 467 US at 843–44.
• 152See US Const Amend V, Amend XIV.
• 153Consider Abbe R. Gluck, What 30 Years of Chevron Teach Us About the Rest of Statutory Interpretation, 83 Fordham L Rev 607, 630 (2014) (“Agencies are Congress’s immediate, frequent, and ongoing statutory interpreters. Because courts are involved, if at all, much more rarely and usually further down the line, they are not [ ] typically on Congress[’s] radar to the extent that courts seem to expect.”).
• 154See text accompanying note 98.
• 155Shifting toward greater focus on predictability would also resemble scholarly proposals to increase consistency in clarity findings by shifting from internal to external assessments of ambiguity—that is, by asking interpreters to predict the likely views of third-parties, rather than assessing their own degree of confidence. See Farnsworth, Guzior, and Malani, 2 J Legal Analysis at 276 (cited in note 9) (suggesting that clarity judgments “can be disciplined, and the effect of policy preferences avoided, by using an external inquiry rather than an internal one”).
• 156While circuit splits are not nearly as salient in Chevron case law as compared with AEDPA and qualified immunity, there is some support for considering them. See Smiley v Citibank (South Dakota), NA, 517 US 735, 739 (1996). See also de Osorio v Mayorkas, 695 F3d 1003, 1016 n 1 (9th Cir 2012) (en banc) (M. Smith dissenting) (noting a circuit split on the relevance of circuit splits to Chevron).
• 157See text accompanying note 81.
• 158For suggestions that Chevron might apply differently in the Supreme Court as opposed to lower courts, see Brand X, 545 US at 1003 (Stevens concurring); Michael Coenen and Seth Davis, Minor Courts, Major Questions, 70 Vand L Rev 777, 781–82 (2017). See also note 121.
• 159See Baude and Doerfler, 117 Mich L Rev at 327–28 (cited in note 60).
• 160Posner and Vermeule allow that a judge might discount the votes of other judges, but only if “the first judge disagrees with the reasons of the other judges” and if “the first judge believes that the spurious reasons provided by the other judges actually motivated the other judges’ votes.” Posner and Vermeule, 105 Georgetown L J at 181 (cited in note 34). As an example, the authors imagine a judge who credibly explains that President Vladimir Putin directed his vote. Id. The authors assert that their dual conditions are “rarely satisfied” because judges’ expressed reasons generally do not track their true motivations. Id at 181–82. But even if so, additional discounting seems justifiable. For example, if a judge defended his vote based on outlandish reasoning, another judge might plausibly infer that the first judge’s motivation is non-legal or otherwise impermissible—even if not quite as impermissible as having “received a phone call from Vladimir Putin.” Id at 181.
• 161For instance, Posner and Vermeule have suggested that the justices should account for one another’s votes partly based on principles of “epistemic humility.” See id at 163–66. See also note 163.
• 162See generally Baude and Doerfler, 117 Mich L Rev 319 (cited in note 60).
• 163But see Posner and Vermeule, 105 Georgetown L J at 176 (cited in note 34) (positing that if five justices think a statute is clear in one direction, and four think it is clear in another direction, then all should vote that the law is unclear).
• 164See generally Joanna Schwartz, The Case against Qualified Immunity, 93 Notre Dame L Rev 1797 (2018).
• 165Pearson, 555 US at 231, quoting Harlow v Fitzgerald, 457 US 800, 818 (1982).
• 166Mullenix v Luna, 136 S Ct 305, 308 (2015) (“A clearly established right is one that is sufficiently clear that every reasonable official would have understood that what he is doing violates that right.”) (quotation marks and citation omitted). See also Anderson v Creighton, 483 US 635, 640 (1987).
• 167Other commentators have observed the basic similarity between these doctrines. See Litman, 106 Cal L Rev at 1493–94 (cited in note 15); Aziz Huq, Judicial Independence and the Rationing of Constitutional Remedies, 65 Duke L J 1, 13 (2015). One important potential difference is that circuit precedent cannot in itself justify relief under § 2254(d) but might be able to do so for qualified immunity. See note 169.
• 168See Hernández v Mesa, 137 S Ct 2003, 2007 (2017).
• 169Even though many or all executive officials should care about circuit court precedent in at least some situations, the Court has recently and repeatedly reserved whether circuit precedent can create clear law for qualified immunity purposes. See, for example, Reichle v Howards, 566 US 658, 665–66 (2012) (“[a]ssuming, arguendo, that controlling Court of Appeals’ authority could be a dispositive source of clearly established law in the circumstances of this case . . .”). See also Richard M. Re, Should Circuit Precedent Deprive Officers of Qualified Immunity? (Re’s Judicata, Nov 17, 2014), archived at http://perma.cc/2MBF-ZXC9. The Court’s recent reservations on this score may stem from an instinct to make qualified immunity conform to the standards of § 2254(d)(1). See note 98. The Court may also want to avoid the dissonance of treating as “clearly established” a view that the justice themselves regard as incorrect. See note 30.
• 170See, for example, Hernández, 137 S Ct at 2007 (“The qualified immunity analysis thus is limited to the facts that were knowable to the defendant officers at the time they engaged in the conduct in question.”) (quotation marks and citation omitted).
• 171See Wilson, 526 US at 617. An ancillary set of doctrinal principles governs when courts should exercise their acknowledged authority to reach the merits while finding qualified immunity. See, for example, Camreta v Greene, 563 US 692, 707 (2011).
• 172Procunier v Navarette, 434 US 555, 562 (1978).
• 173Further, qualified immunity cases sometimes seem to assume that officials can recall and apply fact-specific case holdings, even during quickly developing situations. See, for example, Manzanares v Roosevelt County Adult Detention Center, 331 F Supp 3d 1260, 1293–95 n 10 (D NM 2018). That approach could put an impossible burden on police, but the dynamic in practice is quite different: because fact-specific holdings are rarely on all fours with new cases, officials can point to the lack of any on-point precedent in order to obtain qualified immunity. A more desirable approach might focus instead on a small number of principles that officials could realistically assimilate and then ask whether the officer reasonably applied those principles. See note 77.
• 174The court’s insistence on “objective” tests in this and other clarity contexts also reflects a qualification of its commitment to prediction. See Anderson, 483 US at 639. Though a court might want to know what this specific officer could have predicted, the objective approach avoids intensive factfinding and fosters predictable decision-making on the issue of predictability itself.
• 175See, for example, Pearson, 555 US at 231; Harlow, 457 US at 807.
• 176Ashcroft v al-Kidd, 563 US 731, 743 (2011), quoting Malley v Briggs, 475 US 335, 341 (1986).
• 177See, for example, al-Kidd, 563 US at 741 (emphasizing that, “[a]t the time” of the officials’ alleged conduct, “not a single judicial opinion had held” that the conduct was unconstitutional). On the possibility of violations that are obvious even without a case on point, see note 187.
• 178See text accompanying note 113.
• 179See Carey v Piphus, 435 US 247, 254 (1978) (suggesting that “the basic purpose of a § 1983 damages award should be to compensate persons for injuries caused by the deprivation of constitutional rights”).
• 180See text accompanying notes 101–12, 119. But efforts to redefine legal clarity are especially promising in connection with qualified immunity because that doctrine lacks any explicit basis in statutory text and has in effect evolved into a common law doctrine. See Ziglar v Abbasi, 137 S Ct 1843, 1872 (Thomas concurring in part and in the judgment), citing Reichle v Howards, 566 US 658, 664 (2012) for the proposition that the “clearly established law” standard reflects a balancing of competing interests.
• 181John C. Jeffries Jr, What’s Wrong with Qualified Immunity?, 62 Fla L Rev 851, 867 (2010).
• 182Id at 868.
• 183Jeffries also suggests that his reformulation of the test is meant “to signal that borderline violations would not trigger damages liability.” Id at 867–68. The idea of a “borderline” violation could reflect predictive worries, insofar as a close call is hard to anticipate. But that idea could instead be cashed out as a way of attending to legal certainty: when a court is sure that the officer’s action is unconstitutional, qualified immunity is to be withheld; in a “borderline” case, a court would be relatively uncertain of the law, and the defendant would accordingly have qualified immunity.
• 184Id at 868.
• 185Jeffries, 62 Fla L Rev at 868 (cited in note 181).
• 186Id at 868–69. Jeffries gives the example of fabricating evidence of child abuse.
• 187See Hope v Pelzer, 536 US 730, 741–43 (2002) (finding a violation of clearly established precedent where there was no case on point but the violation was so “obvious” and “apparent” that the officer had “fair warning” of liability). So, under current doctrine, a violation that no court has yet declared unconstitutional can still transgress clearly established precedent—but only if any competent actor would have realized it.
• 188For an analogous proposal in the AEDPA context, see text accompanying note 102. Of course, the main text’s proposal can be adjusted in various ways. For example, a court might deny qualified immunity when the official faced only modest predictive difficulty and the court is highly certain.
• 189See text accompanying notes 119–22.
• 190See Wilson, 526 US at 618.
• 191See id. Interestingly, the Court has attended to circuit splits even when they arise after the officer’s conduct, on the theory that the split demonstrates the prior underdetermined state of the law. Id.
• 192See id.
• 193See text accompanying notes 119–22.
• 194But see al-Kidd, 563 US at 746 (Kennedy concurring) (suggesting a way in which “national officeholders should be given some deference for qualified immunity purposes”).
• 195See, for example, Edward C. Dawson, Qualified Immunity for Officers’ Reasonable Reliance on Lawyers’ Advice, 110 Nw U L Rev 525, 561 (2016) (“If lawyers’ advice may support the qualified immunity defense, this will incentivize officers to seek advice before acting in uncertain situations.”).
• 196See Malley, 475 US at 346 n 9.
• 197See Wilson, 526 US at 617.
• 198See Messerschmidt v Millender, 565 US 535, 553 (2012):

  [T]he fact that the “officers sought and obtained approval of the warrant application from a superior and a deputy district attorney before submitting it to the Magistrate” provides further support for “the conclusion that an officer could reasonably have believed that the scope of the warrant was supported by probable cause.”

• 199See Richard H. Fallon Jr, et al, Hart and Wechsler’s The Federal Courts and the Federal System 1050 (Foundation 7th ed 2015), citing Jack Goldsmith, The Terror Presidency: Law and Judgment inside the Bush Administration (Norton 2007).
• 200Baude and Doerfler, 117 Mich L Rev at 340–44 (cited in note 60) (“[W]e see no good reason to categorically exclude reasonable nonjudges from the project of peer disagreement.”).
• 201See Anita S. Krishnakumar, Reconsidering Substantive Canons, 84 U Chi L Rev 825, 856 (2017) (showing that lenity and constitutional avoidance were the early Roberts Court’s most often used substantive canons).
• 202See, for example, Hooper v California, 155 US 648, 657 (1895) (“[E]very reasonable construction must be resorted to in order to save a statute from unconstitutionality.”). The doctrine of constitutional avoidance supplies an example of a venerable clarity doctrine, illustrating that such doctrines and their difficulties are not terribly new.
• 203See, for example, Moskal v United States, 498 US 103, 108 (1990) (“[W]e have always reserved lenity for those situations in which a reasonable doubt persists about a statute’s intended scope even after resort to ‘the language and structure, legislative history, and motivating policies’ of the statute.”) (citation omitted).
• 204For a discussion of the plain meaning rule, which is even more central to statutory interpretative practice, see notes 223 and 246 and accompanying text.
• 205See Clark v Martinez, 543 US 371, 381 (2005) (noting that constitutional avoidance is “a tool for choosing between competing plausible interpretations of a statutory text”); United States v Universal C.I.T. Credit Corp, 344 US 218, 221–22 (1952) (noting that, under the lenity doctrine, when a “choice has to be made between two readings of what conduct Congress has made a crime, it is appropriate, before we choose the harsher alternative, to require that Congress should have spoken in language that is clear and definite”).
• 206Barber v Thomas, 560 US 474, 488 (2010) (“[T]he rule of lenity only applies if, after considering text, structure, history, and purpose, there remains a grievous ambiguity or uncertainty in the statute, such that the Court must simply guess as to what Congress intended.”) (quotation marks and citations omitted).
• 207See, for example, Bond v United States, 572 US 844, 855–56 (2014). See also Neal Kumar Katyal and Thomas P. Schmidt, Active Avoidance: The Modern Supreme Court and Legal Change, 128 Harv L Rev 2109, 2127–29 (2015); Re, 18 Green Bag 2d at 415–16 (cited in note 54) (discussing examples of textually adventuresome avoidance holdings).
• 208Compare Charles D. Weisselberg and Su Li, Big Law’s Sixth Amendment: The Rise of Corporate White-Collar Practices in Large U.S. Law Firms, 53 Ariz L Rev 1221, 1224 (2011) (positing a “developing norm that corporate officers and employees ought to be represented in white-collar criminal cases . . . by . . . [counsel] at the nation’s leading corporate law firms, most of whom are former federal prosecutors”), with Heather Baxter, Too Many Clients, Too Little Time: How States Are Forcing Public Defenders to Violate Their Ethical Obligations, 25 Fed Sent Rptr (Vera) 91, 91–92 (2012) (discussing the near impossibility for public defenders to fulfill even basic ethical obligations).
• 209Consider United States v Rodgers, 466 US 475, 484 (1984) (concluding that “the existence of conflicting cases from other Courts of Appeals made review of [the relevant] issue by this Court and decision against the position of the respondent reasonably foreseeable”).
• 210In this regard, consider Justice Oliver Wendell Holmes Jr’s view that “a fair warning should be given to the world in language that the common world will understand, of what the law intends to do if a certain line is passed.” See McBoyle v United States, 283 US 25, 27 (1931) (emphasis added).
• 211Lenity’s concern for notice and fairness—and, therefore, for predictability—sits alongside other legally recognized goals. See generally Sohoni, 62 Duke L J 1169 (cited in note 79) (discussing the New Deal Court’s structural concern for legislative authority). See also generally Intisar A. Rabb, The Islamic Rule of Lenity: Judicial Discretion and Legal Canons, 44 Vand J Transnatl L 1299 (2011) (exploring the interplay between legislative supremacy and individual fairness in both United States and Islamic lenity principles).
• 212Already, the Court seems less interested in lenity when the case at hand involves white collar crime. See Zachary Price, The Rule of Lenity as a Rule of Structure, 72 Fordham L Rev 885, 927–28 (2004).
• 213Consider Blodgett v Holden, 275 US 142, 147–48 (1927) (Holmes concurring) (calling invalidation of a statute on constitutional grounds “the gravest and most delicate duty that this Court is called on to perform”).
• 214Reasoning similarly, Farnsworth, Guzior, and Malani suggest that Justice Antonin Scalia’s espousal of a notice-based view of lenity might explain why he was relatively willing to insist on lenity, and they further suggest that lenity determinations be informed by external assessments of ambiguity. See Farnsworth, Guzior, and Malani, 2 J Legal Analysis at 282–83 (cited in note 9).
• 215See, for example, Edward J. DeBartolo Corp v Florida Gulf Coast Building and Construction Trades Council, 485 US 568, 575 (1988) (explaining that courts will “not lightly assume that Congress intended to infringe constitutionally protected liberties or usurp power constitutionally forbidden it”).
• 216See, for example, Rostker v Goldberg, 453 US 57, 64, 83 (1981). This premise is especially plausible when constitutional principles turn in part on technical matters or current public views. See, for example, Turner Broadcasting System, Inc v FCC, 512 US 622, 665–66 (1994) (Kennedy) (plurality).
• 217A competence-based theory of constitutional avoidance would of course rest on a picture of institutional behavior that could turn out to be wrong. So here as elsewhere, empirical argument outside the four corners of the relevant legal text could play a role in altering the law of clarity and, therefore, whether the text is legally clear.
• 218See Re, 18 Green Bag 2d at 408 (cited in note 54) (discussing an approach that “calls for consideration of non-textual factors”—namely “pragmatism” and “legislative purpose”—“when determining how much clarity is required for a text to be clear”). See also Doerfler, 116 Mich L Rev at 533–36 (cited in note 54) (elaborating a similar approach that differs primarily in being exclusively non-purposivist). While developing a philosophically rich account consistent with the pragmatic aspect of the “New Holy Trinity,” Doerfler rejects its purposivist aspect. Doerfler’s main criticism is that it would be double counting to consider legislative purpose both in determining the applicable clarity threshold and then again when assessing the law’s meaning. But purpose can play different roles in those two stages. To give just one example: when setting the clarity threshold, legislative purpose can establish that seemingly peripheral interests are in fact legally central, such that their abridgement requires a higher confidence level; and then, when assessing the law’s meaning, different purposive considerations can resolve linguistic ambiguity.
• 219This approach resembles some of the potential certainty-based approaches to § 2254(d)(1) and qualified immunity canvassed in earlier Sections. See Parts II.A and II.C.
• 220See Farnsworth, Guzior, and Malani, 2 J Legal Analysis at 282 (cited in note 9) (suggesting that “people should not be imprisoned unless the courts are certain that the legislature intended such a result”). See also Doerfler, 116 Mich L Rev at 568–72 (cited in note 54).
• 221See United States v Bass, 404 US 336, 348 (1971) (noting both “the seriousness of criminal penalties” and “the moral condemnation of the community” before concluding that lenity reflects “the instinctive distaste against men languishing in prison unless the lawmaker has clearly said they should.”), quoting Henry J. Friendly, Mr. Justice Frankfurter and the Reading of Statutes, in Benchmarks 196, 209 (1967).
• 222See Doerfler, 116 Mich L Rev at 571 (cited in note 54) (arguing that lenity’s relative weakness suggests “that courts regard criminal cases as having remarkably low stakes”).
• 223See, for example, King v Burwell, 135 S Ct 2480, 2489 (2015) (“If the statutory language is plain, we must enforce it according to its terms.”).
• 224See Bass, 404 US at 347 (explaining that a court should apply lenity when, “[a]fter ‘seiz[ing] every thing from which aid can be derived,’ we are left with an ambiguous statute”), quoting United States v Fisher, 6 US (2 Cranch) 358, 386 (1805) (Marshall); Shon Hopwood, Clarity in Criminal Law, 54 Am Crim L Rev 695, 698 (2017) (“[T]he Court has, by ranking lenity last in the interpretive process, ‘all but guarantee[d] its irrelevance.’”), quoting Dan M. Kahan, Lenity and Federal Common Law Crimes, 1994 S Ct Rev 345, 386.
• 225See Richard H. Fallon Jr, Implementing the Constitution 76–101 (2001) (discussing doctrinal tests that are used to “implement” the Constitution). See also generally Mitchell N. Berman, Constitutional Decision Rules, 90 Va L Rev 1 (2004).
• 226See generally Jacob E. Gersen and Adrian Vermeule, Chevron as a Voting Rule, 116 Yale L J 676 (2007). Suggesting a very different kind of substitution, Professor Mila Sohoni has argued that the New Deal Court lowered clarity thresholds for vagueness and other doctrines while allowing mens rea requirements to satisfy due process concerns. See generally Sohoni, 62 Duke L J 1169 (cited in note 79). In other words, the Court offset changes in clarity doctrines by altering related constitutional rules.
• 227See Messerschmidt v Millender, 565 US 535, 546–47 n 1 (2012) (reiterating that “‘the same standard of objective reasonableness that we applied in the context of a suppression hearing in Leon defines the qualified immunity accorded an officer’ who obtained or relied on an allegedly invalid warrant”), quoting Malley v Briggs, 475 US 335, 344 (1986); Groh v Ramirez, 540 US 551, 565 n 8 (2004). Courts have also equated clarity for purposes of qualified immunity with liability under a federal statute. See note 67.
• 228See, for example, Groh, 540 US at 563, 565 n 8 (applying reasoning from an exclusionary rule case in a qualified immunity case).
• 229Consider Samuel L. Bray, On Doctrines That Do Many Things, 18 Green Bag 2d 141, 148–50 (2015) (noting the benefits of a limited number of “multi-function doctrines” as compared to many varied “single-function doctrines”).
• 230See Part II.D.
• 231See note 227.
• 232Professor Jennifer Laurin provides the leading criticism of the more general pattern of “convergence” between qualified immunity and the good-faith exception. See Laurin, 111 Colum L Rev at 711–13 (cited in note 15). See also Litman, 106 Cal L Rev at 1493–94 (cited in note 15). Laurin’s analysis raises the interesting possibility that initial efforts to tether these doctrines together might have increased the odds that the Court would merge them in other ways as well. For instance, linking two doctrines’ clarity thresholds at Time 1 might cause courts to merge the doctrines’ recognized goals at Time 2—even if the doctrines did and should have distinct aims.
• 233See Malley, 475 US at 343 (“We do not believe that the Harlow [qualified immunity] standard, which gives ample room for mistaken judgments, will frequently deter an officer from submitting an affidavit when probable cause to make an arrest is present.”); Part II.C (qualified immunity).
• 234Adopting the predictive position of a police officer, courts reject good-faith reliance on a warrant where “no reasonably well trained officer” would rely on it. United States v Leon, 468 US 897, 923 (1984).
• 235See note 228 (noting that Groh found no qualified immunity based on an exclusionary rule case).
• 236See note 138 and accompanying text.
• 237See Part II.B (discussing Kavanaugh, Book Review, 129 Harv L Rev (cited in note 9)).
• 238See Kavanaugh, Book Review, 129 Harv L Rev at 2137–41 (cited in note 9).
• 239See Part II.B (discussing Justice Kavanaugh’s critique). For an example of a particularly rule-like clarity doctrine, see note 40.
• 240Justice Kavanaugh’s critiques are thus far limited to clarity doctrines pertaining to statutory interpretation and so exclude case-law based clarity doctrines associated with § 2254(d)(1) and qualified immunity. Perhaps the latter doctrines are, or should be, less objectionable on Justice Kavanaugh’s view because of their focus on predictability.
• 241See text accompanying note 156.
• 242See, for example, note 12 (noting Judge Raymond Kethledge’s assertion that he has never found statutory ambiguity).
• 243Compare Philip Hamburger, Is Administrative Law Unlawful? 316 (Chicago 2014), with Nicholas R. Bednar and Kristin E. Hickman, Chevron’s Inevitability, 85 Geo Wash L Rev 1392, 1397–98 (2017).
• 244Consider Cass R. Sunstein, Incompletely Theorized Agreements, 108 Harv L Rev 1733, 1735–36 (1995). Take the many canons representing soft nondelegation principles: those precepts could represent a compromise between supporters of strong nondelegation principles and opponents of all nondelegation principles. Consider generally Cass R. Sunstein, Nondelegation Canons, 67 U Chi L Rev 315 (2000).
• 245On different justifications for Chevron, see notes 133–34.
• 246See Frederick Schauer, The Practice and Problems of Plain Meaning: A Response to Aleinikoff and Shaw, 45 Vand L Rev 715, 724–25 (1992). But see William Baude and Ryan D. Doerfler, The (Not So) Plain Meaning Rule, 84 U Chi L Rev 539, 539 (2017) (criticizing the intuition that the plain meaning rule is a desirable “compromise”). Part of Baude and Doerfler’s critique rests on the ambiguity of what is “plain,” or clear. Id at 558–62, citing Kavanaugh, Book Review, 129 Harv L Rev at 2137–38 (cited in note 9) and Farnsworth, Guzior, and Malani, 2 J Legal Analysis at 276 (cited in note 9).
• 247See Kavanaugh, Book Review, 129 Harv L Rev at 2121 (cited in note 9) (expressing the author’s main goal: “To make judges more neutral and impartial in statutory interpretation cases.”).
• 248See Barnett, Boyd, and Walker, 71 Vand L Rev at 1523–25 (cited in note 137). See also Adam M. Samaha, If the Text Is Clear—Lexical Ordering in Statutory Interpretation, 94 Notre Dame L Rev 155, 209 (2018) (reporting an empirical study showing “mixed judicial success” at using clarity rules when lexically ordering legal sources).
• 249Malley, 475 US at 341.
• 250See, for example, Ashcroft v al-Kidd, 563 US 731, 743 (2011) (noting that qualified immunity protects “all but the plainly incompetent or those who knowingly violate the law” and that “[former Attorney General] Ashcroft deserves neither label, not least because eight Court of Appeals judges agreed with his judgment in a case of first impression”) (quotation marks and citation omitted).
• 251See Part II.A (discussing 28 USC § 2254(d)(1)).
• 252See Burt v Titlow, 571 US 12, 20 (2013) (“We will not lightly conclude that a State’s criminal justice system has experienced the ‘extreme malfunction’ for which federal habeas relief is the remedy.”) (quotation marks, citation, and alteration omitted).
• 253See, for example, al-Kidd, 563 US at 743 (after noting that only the “plainly incompetent or those who knowingly violate the law” are unprotected, concluding that “Ashcroft deserves neither label”).
• 254The Court could alternatively calibrate the criticism or sympathy it exhibits for the lower court being reversed. At times, for example, the Court has heaped extra criticism on lower courts that had found clear law in an especially indefensible way. See, for example, Harrington v Richter, 562 US 86, 92 (2011) (opening by drawing attention to the “judicial disregard [ ] inherent in the opinion of the Court of Appeals for the Ninth Circuit here under review”). If combined with statements that sympathize with lower court errors in other cases, that approach could create a state of affairs where only some clarity-based reversals are deemed harsh or stigmatizing.