1. Introduction
This Essay revisits the works of Frederick Schauer that examine the nature of legal rules and when having those rules becomes necessary. Specifically, I focus on and expand Schauer’s justification for the rules of evidence. In Part II, I present Schauer’s fundamental idea that establishes the necessity of rules in the domains of morality, politics, and law while responding to the challenge of act-utilitarianism. In addressing that challenge, Schauer famously singled out the principal-agent settings featuring asymmetrically informed rule-maker and rule-applier and then demonstrated that rules have the ability to minimize this asymmetry—a social benefit that offsets the costs incurred by the over- and under-inclusiveness of the rules. Part III transitions to legal epistemology. There, I examine how Schauer expands the “necessity of rules” idea to the law of evidence. Part IV connects this idea to the reasons that can and cannot justify factual findings in adjudication. A short Conclusion follows.
2. Schauer on the Necessity of Rules
Fred Schauer made multiple contributions to legal theory and general philosophy.Footnote 1 Of these contributions, his fundamental insight, describable as the “necessity of rules,” stands out as probably most significant.Footnote 2
Rules seem to collapse under the pressure of act-utilitarianism. Under act-utilitarianism, any specific rule is extensionally equivalent to its general goal.Footnote 3 For that reason, a direct statement of the goal (G) that a given rule seeks to attain by saying “Do X” will arguably do better than the rule. This is because the directive “Do X” can be substituted by “Do what is rational for attaining G” and because, under some circumstances, doing X may not be sufficient or, alternatively, not necessary for attaining G. In cases of insufficiency, the rule “Do X” will be underinclusive and therefore ineffectual; in cases of redundancy, the rule will be wasteful.
Consider a simple example in which someone tells me “Follow rule R, according to which vehicles should be driven at a speed not exceeding 45mph because doing so would allow you to achieve safety on the road, an important welfare-enhancing goal G.” To this I will object: “This may well be right, but, if so, why not follow a simple prescription “Do what is rational to achieve G, that is, drive safely?"” This objection is consequential because the “45 mph” rule is either overinclusive or underinclusive, or both. For example, the directive “drive a vehicle at a speed not exceeding 45 mph” is underinclusive when more needs to be done to maintain safety on the road; and it comes out overinclusive when driving a vehicle at a speed greater than 45 mph promotes safety better than driving a vehicle at a lower speed. An act-utilitarian prescription “drive safely” consequently can justify, under appropriate circumstances, taking greater precautions than not speeding, as well as doing 60 on a 45. The general prescription “drive safely” thus seems to outperform any particularistic rule geared toward maintaining safety on the road.
Before Schauer, the “extensional equivalence” theory and the prevalence of act-utilitarianism were common wisdom. The only plausible exception to that wisdom was a rule-utilitarian move, developed by John Rawls, in relation to socially beneficial practices. Footnote 4 For example, a rule obligating a person to deliver upon their promise—even when doing so is detrimental to welfare under the given circumstances—still has a virtue of supporting the practice of promise-keeping and making contracts.Footnote 5 This rule-utilitarian consequence, however, still cannot justify rules unrelated to general social practices. The number of socially beneficial practices is limited. The multitude of rules that people have developed and follow over time vastly outnumber those practices, and for that reason, additional justification for rule-utilitarianism was required.
In his celebrated book, Playing by the Rules, Schauer did exactly this. He took the justification of rules a step further by demonstrating their beneficial effect in the principal-agent situations—situations where the rule-maker and the rule-applier are not the same person or institution. This I consider to be one of Schauer’s greatest insights.Footnote 6
Schauer demonstrated that rules and rule-following generate important social benefits within the relationship between those who make rules and those who apply them. This is because there is no guarantee of a symmetry between the rule-maker’s and the rule-applier’s understandings of the nature of the goal that the rule-maker chose to pursue and of the means that the rule-maker deems appropriate for attaining the chosen goal. The rule-maker knows what they want,Footnote 7 but whether the applier of the rules is aware of the rule-maker’s plans and ideas is an open question.Footnote 8 Telling the rule-applier “Act to achieve goal G” is a far cry from securing the alignment between the applier’s decisions and actions and the rule-maker’s vision of the goal and the means for attaining it. In my example, the rule-maker’s vision of safety on the road may never allow driving vehicles at a speed higher than 45 mph, while the applier of the rule may allow such driving when it considers it to be safe enough to avoid accidents.
Under such circumstances, given that the rule-maker knows what the rule-applier does not, the alignment between the applier’s decisions and what the rule-maker wants to accomplish and how is a valuable social consequence. The rules’ overinclusiveness and underinclusiveness still impose a cost on whoever is affected by them, but this cost is justified in each and every case in which the alignment between the rule-maker’s will and the rule-applier’s decisions is critical.
This powerful insight immediately translates itself into adjudication—a procedure for vindicating legal rights and enforcing legal liabilities and obligations. If we want to have a legal system that takes legal rules—and the corresponding rights, liabilities and obligations—seriously, we need to secure the alignment between the lawmakers’ will and the law-appliers’ adjudicative decisions. In securing this alignment, rules—functioning, in Schauer’s words, as entrenched generalizations Footnote 9—perform way better than direct references to the rule-makers’ general goals. Lack of symmetry between the lawmaker’s will and the will’s understanding and implementation by those who apply the law is a viable scenario. Under this principal-agent scenario, the “extensional equivalence” theory breaks down and the lawmaker’s will does not get realized. Setting up particularized rules secures the realization of the lawmaker’s will and improves individuals’ compliance with the law. This social benefit, according to Schauer, offsets the social cost of over- and under-inclusiveness of the rules.Footnote 10
3. Does Schauer’s Theory Work in Factfinding?
The foregoing analysis appears impeccable when we focus upon rules that are the product of the rule-maker’s cost-benefit analysis. We can use this analysis to justify a broad array of legal rules as diverse as the “consideration” requirement for contract formation, the definition of “larceny” in criminal law, and the exceptions to the rule against perpetuities in property law. This analysis still leaves unresolved the big question: Can we use it to justify rules of evidence?
This question focuses upon quintessential, or truth-driven, rules of evidence, that include the suppression of hearsay statements,Footnote 11 corroboration requirements for criminal convictions,Footnote 12 and the best evidence rule.Footnote 13 These rules are geared toward ascertainment of the facts upon which judges and juries decide cases, as opposed to the rules extraneous to factfinding, such as evidentiary privileges that sacrifice the pursuit of the truth in adjudication to protect other interests that include privacy, marital harmony, state secrets, as well as communications between attorneys and clients and patients and therapists.Footnote 14 I will call the quintessential rules of evidence “epistemic rules” because they have an epistemic ambition to get the factfinders to the truth.
Arguably, epistemic rules are unnecessary, because in matters of epistemics—as opposed to cost-benefit decisions—there is little potential for misalignment among different people and institutions. According to this argument, the principal-agent problem is not as serious in matters of epistemics as it is in matters of morality and politics. People can rationally disagree about cost-benefit decisions, but there seems to be no room for rational disagreements about empirical facts. When I say to you that such and such facts are present, and you respond “No, they do not,” one of us is right and the other is wrong.
Based on that simple observation, Bentham and his present-day followers have argued that the law of evidence needs no rules at all; all it needs is a simple prescription for factfinders: “Admit and consider all relevant evidence and form your best judgment as to what happened by using general experience, logic and probability.”Footnote 15 This argument received refinement and support from the Oxford philosopher of science, Jonathan Cohen, who wrote a powerful article on Free Proof,Footnote 16 arguing that cognitive competence of factfinders makes epistemically driven evidentiary rules obsolete—an insight that gained widespread agreement among evidence scholars.Footnote 17
If so, does Schauer’s thesis about the necessity of rules fall apart when it comes to evidence and factfinding? Schauer did not address this question in his acclaimed monograph Playing by the Rules—at least not explicitly. This question consequently remained open for several years. I thought about this question when I wrote my book on the foundations of the law of evidence,Footnote 18 but could not develop a clear answer to it. For me, developing this answer did not seem necessary because I claimed that the so-called “epistemic” rules of evidence are not really epistemic: that’s because all of them allocate the risk of error to one party or another and consequently involve a cost-benefit analysis that can be moral, economic, or both.Footnote 19
The question remained open till 2006. In that year, two things happened. First, the University of Pennsylvania Law Review published Fred Schauer’s article titled “On the Supposed Jury-Dependence of Evidence Law.”Footnote 20 Second, a Harvard-based Climenko Fellow, Charles Barzun, presently a law professor at UVA, wrote and circulated his job-talk paper, Rules of Weight, that subsequently appeared in the Notre Dame Law Review.Footnote 21 Barzun’s paper was profoundly influenced by Schauer,Footnote 22 who was back then stationed at Harvard as well. This paper challenged the conventional dichotomy separating rulified court decisions on the admissibility of evidence from factfinders’ discretionary assignment of weight to admissible evidence.Footnote 23 According to Barzun, lawmakers will do well to lay down rules that accord predetermined weight to particular categories of evidence.Footnote 24 Structuring factfinders’ discretion in this way—he argued—will enhance the overall accuracy of their decisions,Footnote 25 as well as promote fairness.Footnote 26
Schauer’s article on the rules of evidence made a very ambitious claim alongside other claims that I consider less ambitious. Among the less ambitious claims was this: If we take seriously the findings of Amos Tversky and Daniel Kahneman about bounded rationality,Footnote 27 then demanding factfinders to follow epistemic rules—ones that tell what decisions are epistemically correct and incorrect—turns out to be a necessity. Under the bounded-rationality assumption, the case for having factfinding rules aimed at overcoming the decision-makers’ cognitive shortcomings becomes overwhelming. For example, once it is established that factfinders systematically ignore base rates, it is epistemically justified to require enhanced evidential support whenever factfinders certify the occurrence of an unlikely event.
For those who remained skeptical about Tversky and Kahneman’s portrayal of people as probabilistically challenged,Footnote 28 this justification of epistemic rules was still too contingent to defy the act-utilitarian case-by-case approach to factfinding. Schauer, however, made a far more ambitious claim that proceeds from the premise that factfinders are epistemically rational. This claim is more ambitious because it answers a very difficult question: Why care about epistemic rules once we assume cognitive competence and epistemic rationality across the board? Based on this assumption, how can factfinders go wrong in making case-by-case decisions in the absence of rules?
Starting from the epistemic-rationality premise, Schauer argued that epistemically motivated evidentiary rules would still have the advantage over case-by-case reasoning. According to him, epistemic rules can enhance accuracy in the long run of cases, as well as create uniformity and consistency among the reasons that justify factual findings. This claim relied on the literature that identified the advantage of actuarial over clinical decision-making.Footnote 29 According to this literature, under factual uncertainty and the ever-present possibility of human error, actuarial decisions outperform case-specific clinical decisions in terms of overall accuracy in the long run of cases.Footnote 30 If so, then all that needs to be done is to formalize the right actuarial decisions by writing them into rules of evidence. Under Schauer’s terminology, such evidentiary rules will function as “entrenched generalizations.”Footnote 31 According to him, entrenched generalizations are good not only for cost-benefit decisions. Epistemics are also in need of such generalizations, and hence the justification for rulifying the law of evidence.Footnote 32
To illustrate this important argument, consider the corroboration requirement for accomplice testimony that applies in criminal trials.Footnote 33 This evidentiary rule holds that incriminating testimony coming from the defendant’s accomplice requires corroboration, in the absence of which the jury should acquit the defendant.Footnote 34 Assume that the lawmakers opt for the non-discretionary definition of “corroboration,” according to which: (1) corroboration must be extraneous: it must come from evidence other than the accomplice testimony; (2) corroboration must relate to a material issue in the factual dispute between the defendant and the accomplice; and (3) it must negate the defendant’s story and, correspondingly, support the accomplice’s account of the events.Footnote 35 Compare this rule’s application across the board with a rule-free clinical evaluation of accomplices’ testimony as trustworthy or not trustworthy beyond a reasonable doubt. Under the presence of uncertainty and human error, this comparison leads to a number of conclusions. First, the corroboration requirement prevents factfinders from over-valuing the credibility of criminal defendants’ accomplices testifying for the prosecution. Second, under this requirement, the presence of corroborative evidence gives factfinders a noncontingent uniform reason for treating the accomplice’s testimony as trustworthy (unless they didn’t believe that testimony in the first place). Third, and perhaps most importantly, this requirement forestalls a normalization of wrongful convictions.Footnote 36 To see why, consider one hundred cases in which, instead of following the non-discretionary corroboration requirement, factfinders convict criminal defendants based on the accomplice’s testimony that impressed them as being trustworthy. This rule-free framework of decision-making guarantees that factfinders convict one innocent defendant or more and, consequently, normalizes the conviction of the innocent. By contrast, under the corroboration requirement, innocent defendants can be convicted only accidentally and on rare occasions, rather than by design.Footnote 37
Critically, in the context of the principal-agent relationship between the maker and the applier of epistemic norms, the requirement that innocents can only be convicted by accident is best realized via rules. There is no way to secure the implementation of this requirement on a case-by-case basis in a rule-free environment. Consequently, Schauer’s “necessity of rules” theory remains robust in the area of evidence. Application of this theory would minimize the incidence of errors in factfinding when the entrenched generalizations formulated into rules of evidence are directed toward that end.
Schauer summarized this important point in a sequel to his 2006 article, “On Supposed Jury-Dependence of Evidence Law,” that appeared in the anthology on the Philosophical Foundations of Evidence Law Footnote 38:
Evidentiary rules, and not only exclusionary rules, compensate for epistemic deficiencies just as ordinary rules compensate for deficiencies in morality or rationality on the part of those whom they regulate. And as with ordinary behavior-regulating primary rules, such an approach does not come without costs. Following ordinary legal rules may produce suboptimal decisions when those decisions are viewed in isolation, but the degree of aggregate suboptimality is commonly thought less than the degree of aggregate suboptimality that would likely exist were decidedly suboptimal decision makers empowered to make their own determinations of the best all-things-considered decision. Similarly, a rule-based approach to evidence may produce frequent epistemic suboptimalities when it excludes genuinely probative evidence, as Bentham so memorably insisted. But, analogously, the suboptimality of such decisions, even when aggregated, may be less than the suboptimality, in the aggregate, of fact-finding by decidedly suboptimal decision-makers, whether they be judges or members of a jury.Footnote 39
4. Rules as Justifying Reasons
In this part of the Essay, I will examine whether epistemic rules can function not only as accuracy enhancers across the board, as claimed by Schauer, but also as justifying reasons for factual findings in courts of law. In Playing by the Rules, Schauer analyzed the entrenched generalizations, formalized into rules, as justifications for decisions, but he did not extend this analysis into the law of evidence. In what follows, I will provide this extension along the lines drawn by Schauer.
The basic premise for drawing those lines is uncontroversial. All adjudicative decisions—on the issues of law and fact alike—need to be justified. If so, what does it mean for factual findings in courts of law to be justified rather than merely accurate under the given probability threshold (preponderance, clear and convincing, or beyond a reasonable doubt)?
Here, the famous Gettier problemFootnote 40 calls for a viable solution because everyone would agree that a factual finding cannot be justified when factfinders determine it accidentally and not upon reasons that are epistemically correct.Footnote 41 When I see a clock on the wall showing 2 at 2 o’clock, I form a justified true belief that it’s now 2 o’clock. Yet, my justification is not good enough and can be Gettierized in philosophical jargon. My justification is not good enough because, unbeknownst to me, the clock on the wall may be broken, and even a broken clock displays the right time twice a day.
Things become markedly different when I visit London and see the Big Ben clock showing 2. Here, my belief that it’s now 2 o’clock is justified not only by “2” that I see on the clock. This belief is also—and, indeed, primarily—justified because I know, as millions of people do, that this specific clock is taken very good care of and is always (or virtually always) accurate. That is, my seeing 2 on the Big Ben clock is corroborated. This corroboration eliminates the possible disconnect between the truth of my belief and the belief’s justification. My belief that it’s now 2 o’clock is factually correct not only because it happens to be 2 o’clock, as in the “clock on the wall” example. This belief is factually correct for reasons that—unlike the sheer display of “2” on a clock—cannot be accidental. In Alvin Goldman’s terms, this belief was brought about by its actual or virtual truth: the display of “2” on a clock that—based upon longtime experience—never, or virtually never, errs.Footnote 42 Put differently, as did Robert Nozick, my belief that it is now 2 o’clock is connected to, or “tracks,” its truth across virtually all possible worlds.Footnote 43 Had it not been 2 o’clock, I would not have believed that it’s now 2 o’clock because my evidence—the time display on the Big Ben clock, combined with the common knowledge about the clock’s accuracy—would not have prompted that belief.Footnote 44
This discussion gives rise to the following general observation: There must be some sensitivity connection between adjudicative findings of fact and the reasons justifying those findings.Footnote 45 Specifically, reasons supporting those findings should not be in existence when those findings are untrue. When factfinding is rule-free and depends on the factfinders’ evaluations of the evidence, making sure that factfinders comply with this requirement and that sensitivity is present across the board is well-nigh impossible. Setting up sensitivity-driven rules of evidence consequently turns out to be necessary for securing—systemically—that findings made by the triers of fact are justified. Because both judges and jurors may err, such rules should apply in jury and bench trials alike.
The corroboration requirement for accomplice testimony illustrates this point as well. When factfinding is rule-free and the defendant’s fate depends on whether the factfinders believe the defendant’s accomplice “beyond a reasonable doubt,” the factfinders are prone to make three different mistakes. Two of those mistakes—false positives and false negatives—are well known. Factfinders may convict an innocent defendant for wrong reasons, and they may also acquit a guilty criminal when—rightly or wrongly—they believe that evidence pointing to his guilt does not eliminate all reasonable doubts. The third type of mistake may be less familiar, but is nearly as pernicious as the first. Factfinders may convict a factually guilty defendant for epistemically wrong reasons. Conviction of a guilty defendant for reasons that are epistemically wrong and are bound to be repeated is almost as bad as conviction of the innocent. Such verdicts lead to conviction of innocent defendants down the road. The corroboration requirement for accomplice testimony works to prevent both mistakes. Satisfaction of this requirement not only increases the probability of the defendant’s guilt by providing the assurance that the testimony given by the defendant’s accomplice is true. The corroboration requirement requires that factfinders do not believe the accomplice unless they receive corroborative evidence—evidence that would normally not be in existence under the (counterfactual) scenario in which the accomplice gives false testimony.
Such an outcome can only be achieved with the aid of rules. Requiring individual factfinders to follow a “sensitivity” standard is less effective than formulating rules that erect formal barriers to conviction, such as the suppression of hearsay statements for lack of cross-examination and the corroboration requirements for accomplice testimony and for a defendant’s out-of-court confessions. A “sensitivity” standard, designed by the lawmaker to govern a broad range of cases, is unlikely to be applied consistently by factfinders instructed only to treat it as a general precept. Such an instruction is too vague to ensure that the lawmaker achieves its aims across the board. Factfinders applying it may demand either too little or too much from the prosecution: in some cases they will rest content with evidence too weak to justify a guilty verdict, while in others they will return unwarranted acquittals of the guilty. The principal-agent problem identified by Schauer looms large here as well.
5. Conclusion
This short Essay cannot do full justice to the richness and multitude of insights developed by Schauer in Playing by the Rules and elsewhere, nor did it attempt to do so. My goal here was to identify a single important insight that highlights the many scholarly virtues of this great legal philosopher. The “necessity of rules” thesis that cuts across many areas of the law—including evidence—is one such insight. This thesis is a product of Fred’s beautiful mind as a creative and rigorous analytical thinker and a true grandmaster of interdisciplinarity.
Open access
