I. INTRODUCTION
Do people have procedural rights, that is, rights that apply in the process of determining their legal disputes, and, if they do, how do those rights relate to accuracy in factual determination? It is commonly claimed, or assumed, that people do have such rights: a list might begin with the right to an unbiased decision-maker and the right to present evidence and argument. In penal proceedings, the list might also include the right to be presumed innocent and the right against multiple jeopardy. Yet the claim, or assumption, that people have procedural rights has not lacked critics. Those who think that law in general, including its procedural rules, can be fully understood as an instrument for promoting some extrinsic value are inclined to doubt that there are any procedural rights except to the extent that such rights would serve that extrinsic value, and so tend to emphasize factual accuracy.Footnote 1 Others, while not necessarily taking an instrumental approach to law in general, think that the moral arguments in favor of procedural rights are weak.Footnote 2 Against such views, I will defend the proposition that people do have procedural rights in the sense that states are under a duty of political morality to provide those who are subject to their laws with fair procedures for settling disputes about the application of those laws. Entitlements to procedural rules such as fairness and finality are rights in the sense that the state is obliged to provide them to the persons under its jurisdiction simply because they are persons and not because it would serve any other purpose to do so. This obligation flows from the state's duty to treat each person as a free and equal member of the legal order.
Yet there is no escaping the fact that sedulous adherence to procedural rights is capable of impeding accuracy in fact-finding.Footnote 3 And even if one rejects an instrumental view of law in favor of a rights-based view, inaccuracy in fact-finding is a serious problem. From a rights-based point of view, an inaccurate outcome (e.g., a wrongful conviction or an incorrect dismissal of a meritorious cause of action) is itself an instance of injustice, or at least an instance of incongruence between the result the law requires and the result the legal system produces.Footnote 4 Put another way, people's substantive rights are not well protected if the procedures for determining disputes about those rights do not generate reasonably accurate results. So the state also has a duty to provide a reasonable degree of accuracy in fact-finding. Anything the legal order can do to improve the accuracy of fact-finding, within the constraints imposed by whatever procedural rights people have, is therefore to be welcomed. Nevertheless, the duty to provide reasonably accurate procedures is subordinate to the duty to provide fair procedures. That is because the settlement of disputes between free persons, and between free persons and the state, must be conducted in a manner that respects their status as free persons.
II. THE RIGHT TO PROCEDURAL FAIRNESS
A. The Legal Order
In the course of his critique of the right to procedural fairness, Wellman remarks that “the existence of procedural rights is more often presumed than defended.”Footnote 5 This remark is puzzling, as there are many accounts of the nature of the legal order that include the requirement of procedural fairness as an essential element. What these accounts have in common is a commitment to the intrinsic worth of the individual person, sometimes expressed as a commitment to human dignity, autonomy, or freedom. They attempt to show the place of procedural fairness in a legal order that constitutes, or at least places a high value on, human freedom and dignity.Footnote 6 My justification for procedural rights is based on this tradition, specifically, on an account of law that treats each individual person as a free, purposive, and choosing agent who interacts with many other such agents. All of these agents need to live together in a way that is consistent with the freedom and purposiveness of each of them. Each person wants to use whatever means are available to him or her to pursue whatever purposes he or she chooses, but in order that no one person infringes the freedom and purposiveness of others, each must be constrained by law, that is, by a set of rules and principles that ensures mutual consistency in the use of means. As long as everyone structures their conduct in accordance with the law, each is free in the sense that they are not subject to the choices of others but only to the public and impersonal requirement of compliance with the law. And so the law itself must be established and enforced by a public authority that is understood to act on behalf of all, not by any private individual or group; otherwise, all the others would be subject to the choices of that private individual or group.Footnote 7 In short, it is the task of the state to establish and enforce a legal order.
The rules established by the state will therefore define what means each person has: they will include general norms concerning property rights, permissible and impermissible uses of property, the formation and discharge of voluntarily created rights and duties, and the governance of relationships such as employer and employee or parent and child—in short, private law. They will also include the rules and principles that constitute the state, that assign the state tasks appropriate to it, and that govern the relationship between the state and those subject to its authority—in short, public law. They may also include rules and principles concerning punishment of people for disobeying other rules—in short, penal law. But in this picture, the task of the state is principally (perhaps solely) to ensure the freedom of each and the consistency of that freedom with the freedom of all by defining and enforcing the rules of public and private law. Rather than being instrumental to any other value, the legal order constitutes freedom.Footnote 8
Now, you might not accept this general picture of the law; you might think that the legal order is best understood not as constituting freedom (as this picture proposes) but as creating an intrinsically valuable state of affairs (as this picture is sometimes understood). This kind of view is, I think, shared by commentators who think that a legal order that constitutes freedom in the sense just described is not an end in itself but is an essential instrument for achieving something that is good for humans but that is at least in principle distinct from the legal order itself. On this kind of view, the ends pursued by law can be as diverse as enhancing human dignity,Footnote 9 supporting the capabilities of each,Footnote 10 or promoting economic development.Footnote 11 On this kind of approach, although law's constitution of freedom is not an end in itself, it is so closely connected with the end that law does ultimately aim to achieve that it is difficult to treat law as merely instrumental, in somewhat the same way that it is difficult to understand breathing as merely instrumental to the oxygenation of the blood. Or, in an even greater departure from the picture that I have sketched, one might think that law is an instrument that can be used to pursue ends of any kind, good, bad, or indifferent; yet, even on that view, the choice of law as the instrument carries with it a set of commitments that will look a lot like those generated by the constitutive account.Footnote 12 Sometimes, you have to act as though a set of instruments (e.g., the rules of a game) is not merely instrumental for it to serve as a proper instrument to its purpose (e.g., the pleasure generated by the game). If that is your view of law, my argument for our entitlement to fair and rational procedures will still go through because it depends less on accepting the picture of law as having no purpose other than to constitute something than on accepting that it does, for whatever reason, constitute something of value.
So a legal order consists of authoritatively enacted general norms that structure and govern human behavior and of institutions for administering and applying those norms. In any legal order, there can be disputes about the application of the law's general norms to particular situations. Those disputes can concern the validity of the norm, the meaning or content of the norm, the way the norm applies to particular facts, or the facts to which the norm is to be applied. So every legal order must have a way of determining these disputes. If there is no such mechanism, then in cases of conflict and dispute, there will be no way that each can know what his or her means are and so no way of ensuring the mutual consistency of everyone's actions. Everyone's rights to use their means in accordance with the law must therefore include access to a procedure for resolving disputes about those rights. The procedure must be public in two senses. First, it must be publicly provided because only an institution that can be understood as acting on behalf of all can have the authority to impose a resolution of the dispute on the parties. This requirement of publicness excludes neither settlement nor private arbitration (though it might justify limits on those procedures), but it does exclude arbitration or settlement unsupported by public procedures defining the conditions under which private dispute resolution is possible and permissible and for enforcing arbitration awards and settlements, for otherwise those resolutions would be no more than private and unenforceable recommendations. Second, the outcome and its supporting reasoning must be publicly observable, so that others can be guided by the authoritative determination of the specific right in question (particularly in rem rights) and can take into account the reasoning that supported the tribunal's determination in their own use of their means.
B. Tribunals and Fair Procedures
The usual mechanism for dispute settlement is a tribunal, that is, an institution whose task is to decide disputes about the application of the norms of the legal order.
It is conceivable that a society could have general norms of behavior and yet lack tribunals in this sense because departures from those norms would be dealt with informally.Footnote 13 I doubt whether such a society could be said to be ordered by law because it would have no systematic method of ensuring that its general norms were enforced and applied in cases of disagreement about their content and application. I see no way of excluding, empirically, the possibility that such a society could nevertheless function well, though it seems so unlikely to me as not to be a plausible basis for the claim that we have no procedural rights; in any event, I will not be concerned with that possibility here.
A tribunal, like any other institution, must have rules and principles that constitute it and enable it to serve its function. The particular kind of rules germane to a tribunal are rules of procedure and evidence, that is, ways of bringing disputes to the tribunal, ways of processing those disputes, ways of deciding what information to take into account when deciding them, and ways of deciding how that information can be used. The legal order must also have procedures for enforcing the tribunal's decision, though I will have nothing to say about those procedures in this paper.
The idea of procedure (and therefore of procedural rights) presupposes a distinction between procedure and substance. I will not have a great deal to say about that distinction here, and nothing in my argument will turn on it. I will use the word “procedure” in roughly the same way that lawyers do when we speak of “criminal procedure” or “civil procedure”: the means by which a legal dispute is brought to a tribunal and resolved by the tribunal, regardless of the substantive law in issue. Procedure in this sense includes, among other matters, the rules governing the steps a person needs to take to bring their complaint to a tribunal and that other persons may take to respond to the complaint, the identity of the persons who are allowed to participate, the timelines for dealing with a dispute, and the way the tribunal processes the dispute. Procedure in this sense also includes rules or principles of evidence. The idea of evidence presupposes a distinction between information that can be considered and information that cannot be considered in reaching a decision; it often, though perhaps not necessarily, also includes criteria for the quality of the information in supporting a decision. So procedure encompasses rules governing the form of a complaint and a response to a complaint, the conduct of a hearing, participation in the hearing, and the methods by which the tribunal establishes the facts necessary to substantiate or refute the complaint.
A fair procedure is often said to have something like the following elements:Footnote 14
1. the dispute between the parties should be decided by a tribunal that is and that appears to be impartial;
2. each party should have the opportunity to present the legal and factual basis of its own case and to respond to the legal and factual basis of the opposing party's case;
3. the tribunal should follow its own rules of procedure and evidence; and
4. the tribunal's decision should be reasoned and should be based on the law and on the evidence before it.
These elements, particularly the second one, might seem parochial in that they suggest that only an oral, adversarial, Anglo-American model of procedure could be fair. But that is not so. The same basic elements of procedural fairness can be found in inquisitorial systems and (therefore) in the jurisprudence of the European Court of Human Rights, even though they manifest themselves differently in inquisitorial systems than in the common law jurisdictions.Footnote 15 A fair process does not preclude the tribunal's relying on a dossier assembled in advance and read by the tribunal before any hearing begins, nor does it require any particular rules of evidence, nor does it even require orality. What it does require is impartiality, evenhandedness, and rationality in decision-making. I will therefore take it that these are in general the elements of a fair procedure for resolving disputes about the application of the law's norms to particular situations.
C. The Right to a Fair Procedure
Procedures and procedural rights are inescapably institutional. Like the right to vote or the right “to petition the Government for a redress of grievances,”Footnote 16 procedural rights cannot exist in any positive form apart from the institution (the tribunal) to which they pertain. But that a right is institutional does not entail that it is meaningless apart from the institution to which it pertains; if that were so, then no one could have any rights in connection with any institution. There is no right to vote for one's legislative representatives in the state of nature or in an absolute monarchy because there is no legislature to which that right might pertain, but even in the state of nature, or in a legal order where legislative power is not exercised by a representative body, people have a right to vote in the sense that they have a political right to the establishment of a legislature and to participate in choosing its members. The right to procedural fairness is similar in this respect. There can be no right to procedural fairness in the state of nature because there are no tribunals in the state of nature, but people do have a right that the legal order, once it has constituted a tribunal, provide fair procedures. So, when Wellman says that “there are no preinstitutional procedural rights,”Footnote 17 I understand him not to be making the trivial claim that there are no procedural rights in the absence of institutions but the serious claim that, pre-institutionally, no one has a right that, if institutions are established, the procedures associated with them have any particular content. I think they do. The state has a duty to provide people with tribunals for settling their disputes; since the people whose disputes are being settled are juridically free and equal, the tribunal's procedures must be fair, and since more accurate dispute settlement leads to better protection of rights, the tribunal's procedures must be rational, though it is hard to say whether anyone has a right to any given level of accuracy.
If people have no way of determining their legal disputes, the state does not serve the purpose of ensuring the equal freedom of all. So the right that people have to have their rights defined and enforced (without which their rights mean nothing) implies that the state has a political duty to establish a tribunal. This duty is owed to all the persons who are subject to the law, both the parties to the disputes and those whose might be affected by the resolution of the dispute, because it is their freedom that is at stake in the tribunal's resolution of disputes. So everyone has a right that the state establish a tribunal for the resolution of all their legal disputes.
That the state has a duty to establish a tribunal does not however yet show that the parties to a dispute have a right to a fair procedure. One might think that any procedure for resolving disputes would do. That thought might be right if all that mattered was that disputes be resolved. But the tribunal was established to resolve disputes between persons who are supposed to be free in an appropriate sense, and so the dispute resolution process must be one appropriate for such persons. As it turns out, this requirement implies fair procedures in the law's ordinary understanding of that idea. It also implies that accuracy in fact-finding is an important value in dispute resolution.
Recall the four elements of a fair hearing: the tribunal's being impartial, hearing both sides, following its own rules, and applying the law to the facts it has found. The third and fourth elements are relatively straightforward and might be said to be required by the rule of law even in a very thin sense as it applies to tribunals in particular. Tribunals, like other state officials, are supposed to do the job that is given them and to apply the law that governs that task. So a tribunal should apply the law that applies to it—its rules of procedure and evidence, and it should apply the substantive law to the facts as it finds them. If it did not apply its own procedural rules, or if it applied the substantive law on the basis of some facts other than those it had determined, it would not be doing its assigned task in accordance with the law that applies to it. Everyone whose rights are affected by the conduct of state officials has a right to those officials’ acting lawfully. So everyone has a right to the third and fourth elements of a fair hearing.
The first and second elements, taken together, define the essential fairness of a procedure: the impartiality of the tribunal, in the sense of its neutrality between the parties.Footnote 18 Everyone has a right to a tribunal that is impartial in these ways, regardless of whether the dispute in question is a matter of private law, of public law, or of penal law. But the justification for the requirement of impartiality is not exactly the same in public and penal law as in private law.
As to private law disputes, the requirement of impartiality flows directly from the tribunal's function of deciding disputes between free and equal persons. There is no reason consistent with the equal freedom of the parties for a tribunal to favor one free person over another in its dispute resolution procedures. A tribunal that is biased in favor of one party decides based on who the parties are rather than on the basis of what the law requires of free persons and so undermines the independence of the two private parties from each other by enabling one to impose its choice on the other. Impartiality in turn implies that each party should be able to present its own case and respond to the other; permitting one party to present its case while preventing the other party from doing so would be a kind of unequal treatment inconsistent with the status of each party as a free person independent of the other. Thus, the commonly understood elements of a fair hearing flow directly from the task the state is required to do in dispute resolution.
As to disputes between the state and individuals, the need for impartiality is perhaps not immediately obvious. One might argue that in proceedings between the state and an individual the tribunal should not be impartial but should be biased, perhaps in favor of the state, perhaps in favor of the individual, depending on the nature of the dispute. An argument of this kind would flow from the fact that these disputes are not between two free and equal persons but between one free natural person acting on his or her own interests and an artificial person acting behalf of all. In these disputes, one state official (the tribunal) judges the conduct of another state official (e.g., an administrative decision-maker) in relation to an individual, or judges the truth of the state's allegation of wrongdoing against the individual. So, on the one hand, one might think that the tribunal should be biased in favor of the state because the tribunal, like other state officials, should in general act as an instrument of state policy in resolving the dispute. On the other hand, one might think that because of the imbalance of power between the state and the individual, a tribunal ought to be systematically biased against the state in these disputes, just because the state is an artificial person with no purposes of its own.
It is of course possible to build such biases directly into procedural law. The requirement of proof beyond a reasonable doubt in criminal cases might be characterized as such a bias in favor of the individual; the attenuated procedural rights granted to claimants in many systems of immigration law might be characterized as a bias in favor of the state. But, assuming that is the correct description of such rules, as long as the tribunal applies the rules, as they are, diligently in particular cases, any partiality is due to the law and not to the tribunal's own bias.
The claim that a tribunal ought to favor the state may have some force if the law is understood entirely as an optional instrument for achieving some external good. But if the law is understood as a way of constituting the freedom of individuals (for its own sake or as an indispensable instrument to another value such as human dignity), then they fail because they assume that there is something the tribunal can permissibly do to forward the relevant state policy apart from applying the law diligently. But that is not so. The government, as an artificial person (or as a set of institutions) has no purposes of its own apart from properly administering the law; the individual, whatever their purposes may be, appears before the tribunal solely as a legal subject seeking to be treated in accordance with their rights. Any other purpose the government may have is illegitimate; any other purpose the individual may have is irrelevant. The relevant state policy is embedded in the law. The argument against impartiality would be that the tribunal's task is to forward the policy of the state that underlies whatever public law is being applied.Footnote 19 But from a tribunal's point of view, the only way to forward the state policy embodied in the specific laws under which the particular dispute has arisen is to apply the law correctly to the dispute.
None of this is inconsistent with the claim that the tribunal should interpret the relevant norm in light of its purpose and in light of the individual's status as a free person.Footnote 20 But that interpretive exercise affects only the tribunal's understanding of norm; once that understanding is in place, it gives no warrant for biasing the procedure as between state and individual. That, in turn, means that in both public and private disputes the tribunal should hear from both sides and decide according to the law based on its own procedures and the evidence at its disposal. But those are just the basic requirements of a fair hearing, no different in public law proceedings than in private law proceedings.
It is, on the other hand, sometimes said that criminal proceedings should be biased in favor of the accused person because of the inherent power imbalance between the state and the individual or because of the serious consequences of a prosecution for the individual.Footnote 21 These imbalances and consequences are often real, and they may affect the specification of some of the rules of criminal procedure, but the core entitlements of criminal procedure apply even when they are not present. We do not relax the presumption of innocence or the rule against double jeopardy in specific cases where that kind of rationale does not apply, for example where the accused is fabulously wealthy and the offense minor, in the way that we do relax other rules of evidence and procedure when their rationales do not apply (e.g., case-by-case exceptions to the rule against hearsay, or appellate curing of a first-instance procedural irregularity). On the other hand, when these principles are relaxed, the justifications for relaxing them do not seem to have anything do to with the idea that the tribunal should be biased in favor of the state; rather, it seems to be connected with the need for improved accuracy in adjudication, combined with some argument as to why the relaxation of the principle is not, in context, particularly unfair. For example, the argument that reversing the onus of proof on some offense element would increase the proportion of convictions is never sufficient in itself to justify a departure from the presumption of innocence; there must also be a good argument that reversing the burden of proof on an element is not unfair because, for example, in the particular context the defendant has privileged access to the information relevant to disproving the element in question.Footnote 22
D. Wellman's Critique of Procedural Fairness
Wellman argues that “there are no general, moral, juridical procedural rights.”Footnote 23 The right to procedural fairness that I have just argued for is just such a right. So, it may be helpful to try to understand why Wellman thinks there are no such rights.
Wellman does not make a general argument against the existence of procedural rights. Instead, he argues by way of appeal to intuition from counterexamples. He tries to show that two entitlements often thought to be procedural rights are not: the right against double/multiple jeopardy and the right not to be punished without a (fair) trial. His thought is that everyone takes these procedural entitlements to be at the core of the right to fair procedure, so if these procedural entitlements are not rights, then probably no procedural entitlements deserve to be called rights. But his counterexamples are so weak that they should not unsettle anyone whose intuitions differ from his. I consider his argument against the right to a fair trial here and I consider his argument against the right to finality in Section III.B below.
To show that there is no right to a fair trial, Wellman asks us to imagine the following unusual proceeding. Suppose that someone steals a substantial sum of money from the state treasury. Unable to solve the crime through their usual (fair) procedures, state officials propose to “hold a lottery and … punish the person whose name is drawn.”Footnote 24 The proposal is debated and duly adopted according to the state's democratic procedures. The lottery is held and the person chosen for punishment is Sandra. “[B]y a spectacular coincidence (and unbeknownst to any of the authorities who apprehended and punished her), Sandra is the one who actually stole the money from the treasury.”Footnote 25 Sandra is duly punished and so, it seems, she gets what she deserved. And so, Wellman suggests, her rights were not violated. But the procedure by which she got what she deserved did not in any way comply with the requirements of procedural fairness. And so, Wellman concludes, there is no right against punishment without a fair trial.
This example fails to prove its point because Sandra was not punished; rather, she was scapegoated. Sandra was not punished because she was not chosen by the state to suffer hard treatment for an offense that she was publicly announced to have committed. Wellman recognizes this point but rejects it on the ground that “a person who has been framed by a corrupt authority makes perfect sense when she complains of being wrongly punished.”Footnote 26 That is true, but it does not describe what happened to Sandra. The authorities didn't go to the trouble of framing her, with all of the secrecy and deception that would have required, but instead picked her through an open and transparent process that was deliberately dissociated from anyone's personal responsibility for the offense. A person who is framed is publicly, though corruptly, represented as having committed the crime in question, but “no one seriously believes that [Sandra] is guilty.”Footnote 27 Hard treatment that is neither represented as nor believed to be a public response to criminal wrongdoing is not criminal punishment.Footnote 28
What happened to Sandra was not punishment but scapegoating: she was not chosen for punishment but was picked for hard treatmentFootnote 29 because the community (for some reason) felt a need to impose hard treatment on someone. But was Sandra wronged? Wellman thinks not: he thinks that she got what she deserved for her crime of theft. She was, as it were, scapegoated for her own wrong.Footnote 30 As Wellman would put it, when she committed theft, she forfeited her right against punishment for theft, and so cannot complain that, in the end, she received hard treatment that was equivalent to the punishment she deserved.Footnote 31 It may be that this view can be supported from some private, all-things-considered moral point of view. But from the public point of view of a legal order, Sandra was wronged because she suffered hard treatment for no juridical reason, that is, no reason connected to anyone's rights or to the state's task of creating a legal order that constitutes the freedom of all. Although the procedure by which Sandra was picked was, as stipulated by Wellman, fair, scapegoating itself is wrongful because it is the imposition of hard treatment for no reason related to the scapegoat's conduct.
III. OTHER PROCEDURAL RIGHTS
Everyone has a right to procedural fairness in all kinds of proceedings. But there may be other rights, plausibly characterized as procedural, that go beyond the basic right to procedural fairness. In penal proceedings, such rights would include, I think, the right to be presumed innocent, the right against self-incrimination, the right to reasonable bail pending trial, and the right to representation at trial. Rather than trying to make the case for all of these, I will focus on what is sometimes called the right against double jeopardy or what I will call the right to finality: once a person has been ultimately convicted or ultimately acquitted of a penal offense, he or she has the right not to be charged or tried again for that offense.
A. The Right to Finality
On its face, it is a puzzle how there could be any controversy about the right to finality, as it might be seen as a straightforward, even trivial, implication of the pragmatic need for disputes to be settled, and in a sense it is. But finality is not just a matter of pragmatism. Without finality, no one would ever know what his or her rights were, would never know what means were lawfully available for him or her to act in the world, and would not be able to act with any assurance of avoiding liability. In private law and administrative law, the purpose of dispute settlement is to enable people to know what their rights are so that they can continue to act in the world. And so where the plaintiff and defendant have already settled those questions through litigation, where the plaintiff tries to reopen them, the defendant is entitled to plead a defense such as res judicata. In penal law, finality is as pressing, if not more so, than in private or administrative law. Although penal law typically does not determine anyone's private law rights as such,Footnote 32 it can have a devastating effect on the freedom of the accused. In the modern world, imprisonment and fines are the most basic forms of punishment,Footnote 33 and these forms of punishment directly affect the most basic means that we have for taking actions in the world: personal liberty and economic resources. Even without this consideration, the need for finality in penal proceedings is more urgent, or at least more basic, than the need for finality in civil proceedings. The point of the criminal trial is to determine an allegation of wrongdoing. At the outset of the proceedings, the accused is presumed innocent; without this presumption, everyone would live in a perpetual state of liability to punishment without proof of guilt.Footnote 34 If the presumption is rebutted and the accused is found guilty and so liable for punishment, there is no purpose in prosecuting him or her again. If he or she is acquitted, then for the purposes of penal law he or she must be taken not to have committed the offense; otherwise, he or she lives in a perpetual state of potential liability for an offense for which he or she has already authoritatively been determined not to be liable for. For the same reason that everyone is entitled to be presumed innocent unless and until guilt is authoritatively determined, everyone is entitled to be deemed innocent (for the purposes of the criminal justice system) once authoritatively found not guilty. Thus, the right to assert a defense such as res judicata and the right against double jeopardy are not only entitlements that flow as a by-product of the legal system's need for finality in dispute settlement. They are rights that participants in litigation have in virtue of being free persons.
B. Wellman's Critique of Finality
Against the right to finality, Wellman asks us to consider the following. Suppose it was possible to change the rules of procedure so as to “increase the percentage of guilty people convicted and decrease the number of innocent people wrongly convicted by allowing double jeopardy but increasing the burden of proof that the prosecution must establish.”Footnote 35 Would anyone's rights be violated by this change? Wellman thinks not. Without more specification of the example, neither do I, but without more specification, Wellman's argument does nothing to cast doubt on the justification for the right to finality that I offered above. Despite his recognition of the inescapably institutional nature of procedural rights, Wellman does not explain what he means by “double jeopardy” or what procedural rule he thinks would violate the right against it. For the purposes of discussion, I'm going to assume that he is referring specifically to the American version of the right, and in particular to its effect on appellate procedure.
In the United States, appeals from acquittals are generally not permitted because the presumptive remedy of a new trial is understood to put the accused in jeopardy again, so a person who has been acquitted even in a legally flawed trial benefits from the right to finality.Footnote 36 But other jurisdictions take a different view of this right. In Canada, the prosecution may appeal from acquittal “on any ground of appeal that involves a question of law alone.” This right of appeal was enacted in 1930,Footnote 37 but did not otherwise affect the common law right against double jeopardy.Footnote 38 The constitutional right against double jeopardy (among other procedural rights) was not entrenched until 1982. The constitutional version of the right provides that everyone charged with an offense has the right “if finally acquitted of the offense, not to be tried for it again and, if finally found guilty and punished for the offense, not to be tried or punished for it again.”Footnote 39 It was not long before the Supreme Court of Canada was called on to decide whether the Crown's statutory right of appeal from acquittal violated the accused's constitutional right against double jeopardy. Based on the specific wording of the constitutional text (the word “finally”), but possibly influenced by the long-standing historical status of the Crown's right to appeal, the Court concluded that it did not.Footnote 40 The Crown's right to appeal on a question of law alone does not violate the right against double jeopardy in Canada because that right engages only after the accused has been acquitted in a legally proper trial.
In England, the Crown has no general right of appeal from acquittal; on the other hand, retrial of an acquitted person for certain serious offenses is permitted in very limited circumstances.Footnote 41 The prosecution must apply to the Court of Appeal for a retrial. The Director of Public Prosecutions (DPP) must personally consent to the prosecution's making the application to the court.Footnote 42 Before giving such consent, the DPP must be satisfied that a number of criteria are satisfied, notably including that a retrial would not violate the acquitted person's right against double jeopardy as specified in the European Convention on Human Rights.Footnote 43 To make an order permitting a retrial, the Court of Appeal must be satisfied that (among other things) there is “new and compelling evidence against the acquitted person.”Footnote 44 The prosecution can make only one application for retrial of an acquitted person,Footnote 45 so, if the accused is acquitted at the retrial, the prosecution cannot invoke this procedure again. Thus, in England, a person who has been acquitted of a sufficiently serious offense can under some circumstances be retried for that offense, but only once.
Now, it is likely that these Canadian and English rules would violate the right against double jeopardy as that right is understood in the United States. But it is far less clear that they violate the right to finality as defended above. Each of them can be understood as a different way of concretizing the right to finality in different jurisdictions that, though their common law roots are the same, have different legal histories and traditions. Thus, even if we agree with Wellman's intuition about the permissibility of multiple jeopardy, his argument does not show that there is no right to finality. One might just as well try to show that there is no right to freedom of expression by arguing that there is an intuitive appeal to expanding the right to freedom of expression in some cases (perhaps defamation or breach of confidentiality) while limiting it further in others (perhaps hate speech or pornography). Such arguments invite reflection on the meaning of speech, on the rationales for its protection, and on the question whether there can be justified limits on speech; they do not show that there is no right to free speech. Similarly, Wellman's thought experiment, even if it has intuitive appeal, does not suggest that no one has a right to a final resolution of their dispute with the state. At most, it shows that the U.S. version of the right (assuming that is the one he has in mind) is not the only one imaginable. It invites reflection on whether the rationale for finality extends to the kind of case he envisages and, if it does, how exactly the right against multiple jeopardy should be specified in light of the other features of the legal order and whether a limit on or exception to the right against multiple jeopardy might be justified. But it does not show that there is no right against being repeatedly charged with the same offense; even the English approach does not permit that.
Thus, Wellman gives us no plausible reason to think that we have no general moral right to finality. To the contrary (and puzzlingly in light of his general claim), he provides a very clear articulation of the reasons why finality is morally required. Without a rule against multiple jeopardy, he says, “legal officials could continually harass citizens by repeatedly charging them for the same crime or charging folks when there is very little evidence (since they could simply prosecute again later if the defendant is acquitted). In other words, in the absence of a strict prohibition against multiple jeopardy, there would be too much room for legal officials to abuse their powers, thereby undermining citizens’ capacities to lead good lives.”Footnote 46 This strikes me as an excellent argument in favor of a right to finality. Wellman characterizes it as “instrumental,”Footnote 47 but to me it sounds like a principled justification for a right against multiple jeopardy in a legal order constituted around the idea of human freedom. If it is instrumental, it is so only in the rather attenuated sense of being a contribution to a legal order that is designed to constitute a system of freedom within which people can choose and pursue their own conceptions of the good, not in the strong sense of causally contributing to an independently defined human good.Footnote 48
IV. FAIR PROCEDURES AND RATIONAL FACT-FINDING
Accounts of the importance of procedure sometimes underestimate the tension that procedural rights can create within a dignity-based account of the law. That is because procedural rights can and sometimes do impede the achievement of a substantive outcome that the commitment to human dignity appears to require. If one thinks, plausibly enough, that a robust commitment to human dignity requires the state to define certain wrongs as criminal offences and to prosecute them,Footnote 49 then it is certain that a robust commitment to procedural fairness will sometimes frustrate that requirement because it will predictably result in a number of offenders escaping punishment. Moreover, if one thinks that this commitment extends not just to making efforts to punish offenders but also using the punishment of offenders as an instrument for discouraging other potential offenders, then a robust commitment to procedural fairness will sometimes frustrate that project too. And so one might share a commitment to human dignity or human freedom and nevertheless believe that procedural rights, even the basic demands of procedural fairness, were entirely subordinate to that value: one may believe that the best way to ensure maximum respect for the human interests at stake is to allow tribunals to operate unfairly if, by doing so, they would produce more accurate outcomes in litigation and so perhaps encourage people to comply with the law and to that extent with the demands of human dignity (to the extent that those demands are expressed in the law). On that view, there would be no right to procedural fairness as such, as the decision to grant or withhold that right would depend on a legislative assessment of how likely it was to promote human dignity overall. My argument, in contrast, is that the notion of human dignity, to the extent it can be expressed in law, itself requires certain elements of procedural fairness and that denying the parties these elements is a wrong to them. In that sense, we have a right to procedural fairness. I do not claim that this right is not in tension with other demands that the notion of human dignity may make on the legal order, but I do claim that those tensions can be structured. The central elements of procedural fairness are owed to people involved in disputes about their rights even if respect for those rights may frustrate the achievement of valuable purposes, even the achievement of the substantively correct legal outcome, because persons are entitled to be treated as persons in the settlement of their disputes as much as in any other interaction. But there are many ways for tribunals to go about their task of resolving disputes, and within the constraints imposed by procedural rights, tribunals should be governed by rules of evidence and procedure that tend to promote accuracy (in the sense to be defined below). In these ways, the rights generated by the requirements of procedural fairness are no different from any other rights: they constrain the pursuit of our objectives but do not otherwise specify what our objectives should be or how we should pursue them.Footnote 50
If the facts found by a tribunal were entirely random in the sense that they were uncorrelated with what actually happened,Footnote 51 then the tribunal could not be said to be applying the law to the facts (although the result of any given case might, by accident, be correct). The rule of law, as reflected in the fourth element of procedural fairness, requires something more than random fact-finding. But, on the other hand, perfect fact-finding is impossible; error in fact-finding is unavoidable. Moreover, factual errors can occur whether or not procedures are fair and whether or not the procedures in force are scrupulously adhered to. One might complain that a wrongful result occurred because, although the tribunal conscientiously followed its own rules, the rules themselves are flawed. Perhaps the tribunal's rules tend to exclude reliable evidence (a frequent complaint of those who seek to reform the common lawFootnote 52) or tend to allow reliance on unreliable evidence (a frequent complaint of those who are concerned about the rules governing the admission of an accused person's statementsFootnote 53). Yet even thoroughgoing reforms of the rules of evidence could not eliminate all errors. If, as I have argued, there is a right to procedural fairness, how should that right be understood in light of the importance of accuracy in fact-finding, particularly given that accuracy in fact-finding itself contributes to the protection of rights?
Suppose a decision is substantively incorrect, in that someone is found liable when, if the facts had been correctly determined, he or she would not be (e.g., the conviction of the factually innocent), or that someone is found not liable when, if the facts had been correctly determined, he or she would be (e.g., the acquittal of the factually guilty). These two types of incorrect results are analogous to Type I error (false positives) and Type II error (false negatives) in certain scientific testing procedures. There is no known procedure, either in law or in science, that can completely eliminate both types of error (if there was such a procedure, we would surely use it for many purposes other than determining facts for the purpose of dispute resolution). It is therefore useful to imagine that we can construct what one might call a “procedural loss function,” analogous to the statistician's loss function.Footnote 54 A procedural loss function is the sum of several terms, each being the disvalue of a type of error multiplied by its probability, where the probability of each type of error is in turn determined by the tribunal's rules and procedures. The relative disvalue attached to each type of error is a question of political morality. The weighting of one type of error may be higher than the other, as in the slogan “it is better that ten guilty persons escape, than that one innocent suffer.”Footnote 55 Or the weighting may be different for different kinds of error: perhaps the disvalue of an error (of either kind) is higher in a murder trial or a billion-dollar class action than in a dispute over a parking violation or an individual claim for $50. Given the disvalues, the value of the procedural loss function can be changed by changing procedures and thus altering the probability of each type of error. Other things equal, one would like to minimize the procedural loss function. In the rest of this paper, all references to increases or decreases in the accuracy of fact-finding should be read as shorthand for changes in the value of the procedural loss function that are attributable to changes in a tribunal's rules and procedures rather than changes in the disvalues attached to the various types of error: improvements in accuracy are reductions in its value and deteriorations in accuracy are increases in its value, given the judgment of political morality embedded in the weighting of the errors.
It is very likely that the basic requirements of procedural fairness, on balance, contribute to rather than detract from accuracy in adjudication.Footnote 56 There would be few prosecutions, fewer convictions of the guilty, and little if any criminal punishment if only the defense and not the prosecution were allowed to present its case, and there would be a lot of civil claims and a lot of damage awards unjustified by the facts if defendants were not allowed to respond to plaintiffs’ claims. But it is not possible to go beyond such generalities and quantify the impact of particular changes in the rules of evidence and procedure on the procedural loss function in any serious way. There are at least two reasons for this. First, despite the efforts of social scientists, we have only the vaguest idea of the extent to which various procedural rules contribute to inaccuracy. For example, there is a lot of social-science evidence about the extent to which different interviewing techniques are likely to encourage interviewees to speak truthfully (that is, sincerely),Footnote 57 and we can be fairly confident that police officers would obtain proportionately more sincere and fewer false confessions and witness statements if they restricted themselves to the techniques that have been shown to be better at eliciting sincere statements. So it might be thought that we should adopt rules of evidence that would encourage the use of such interview techniques. But it is very hard to tell whether this would tend to reduce the value of the procedural loss function. It is hard to know whether techniques that tend to produce sincere witness statements also tend to produce more information overall; it is possible that good interviewing techniques tend to produce fewer statements than bad ones, in which case the widespread adoption of good interview techniques would generate less information for fact-finders to consider. Moreover, sincerity is not the only measure of reliability; it might not even be a good one, unless the correlation between sincerity and truth is strong. The fact (if it is a fact) that statements from accused persons obtained through coercive methods of interrogation are likely to be false does not imply that witness statements obtained through noncoercive methods of interrogation are likely to be true. In any event, the precise quantitative impact of alternative interview methods on the outcomes of actual trials is not known. Moreover, a change in a procedural rule is likely to have effects on litigation strategies that will confound any effort to determine its effect on the procedural loss function.Footnote 58 For example, a more generous rule for admissibility of confessions is likely to induce more guilty pleas (regardless of the actual innocence or guilt of the suspects). And so it is very difficult to assess the impact of this type of change on accuracy.
Second, the choice of the weight to be attached to the different types of error is not an empirical question at all: it is a normative question about the appropriate rate at which different errors are traded off.Footnote 59 I have no idea what the procedure for determining those weights would be.Footnote 60 So invoking the procedural loss function is purely a way of talking about how changes in procedural rules affect our assessment of how well the tribunal does its job. It is a reminder that any changes in the rules of procedure or evidence are likely to increase the accuracy of fact-finding in some types of cases while reducing it in others, so that there are always trade-offs in the choice of procedural rules.
What, then, is the relationship between the problem of accuracy in fact-finding and procedural fairness? It is tempting to say that everyone has the right to a minimized procedural loss function, where “everyone” includes not only the parties to litigation but also those who are affected by the outcomes of litigation; surely the rights of all are better protected and the guidance function of the law is improved when the procedural loss function is minimized. Other things equal, that is of course true. But it cannot support a right to accuracy in fact-finding. Pragmatically, to minimize the procedural loss function would require the state to devote infinite resources to dispute resolution.Footnote 61 Structurally, even if we recharacterize the proposed right as a right to a minimized loss function given the resources the state has decided to devote to dispute resolution, such a proposed right would require a court to take into account not only the effects of the particular procedural rule the litigant has chosen to challenge but also the interaction of a change in that rule with all the other rules of the system, most likely on the basis of very inadequate data.
It is, I think, more plausible to say that the likely effect of a change in procedure on the accuracy of fact-finding is relevant to the way specific procedural rules are constructed and interpreted in light of their function in a rights-based legal order. It is relevant because, other things equal, a more accurate system of fact determination is better for the protection of rights. But it cannot be the only factor, for at least two reasons. First, as I have already indicated, the effect of any change in procedural rules on accuracy is likely to be very difficult to predict; the problem is inherently complex and the relevant data, though far more abundant now than in the past, are not going to determine it. Second, any construction or interpretation of a procedural rule must be a construction or interpretation of it in light of the its role in a rights-based legal order. So, even if we knew that abolishing a procedural rule would improve accuracy (which, typically, we don't), it would be impermissible to abolish the rule if that rule was also a procedural right.
Consider, by way of example, the rules governing the taking of statements from suspects and the use of those statements at trial. Consider specifically the common law rule that an involuntarily obtained statement is not admissible against the accused at trial. The common law justification for this rule was that such statements were unreliable; more recent justifications have attempted to explain it as an aspect of the right against self-incrimination.Footnote 62 My view, which I will not expound in detail here, is that the right against self-incrimination flows from the right to be presumed innocent and that, whatever the exact contours of the right against self-incrimination may be, it includes a right against having involuntarily obtained statements put against the accused person at trial. Now, suppose it is proposed that the rule excluding involuntary statements should be abolished on the ground that doing so would increase the accuracy of fact determination. It seems to me that neither the argument for nor the argument against this proposal can depend on its effect on the loss function. Excluding involuntary confessions will certainly reduce the number of wrongful convictions (those based on false confessions), but it will likely increase the number of incorrect acquittals for the simple reason that an involuntary confession is sometimes true. We have no obvious way of measuring either of these effects or of balancing them against each other. Thus, if there is a right to the exclusion of involuntary confessions (or, more generally, of systematically unreliable evidence), that right must be based at least in part on the same values that underlie procedural fairness itself, not on the claim that granting such a right would reduce the value of the procedural loss function. On the other hand, if I am right to base the right to exclusion of an involuntarily obtained statement in the right against self-incrimination, the argument that respecting the right would tend to reduce the accuracy of fact-finding is just beside the point because the exclusion of involuntary statements is not justified by its effects on accuracy but by its role in a system of procedural law that takes the right against self-incrimination seriously.
And yet, having said that, there are many ways in which the proposed right against having involuntary statements put against an accused can be spelled out in detail. For example, I have so far said nothing about what it is that makes a statement involuntary. The common law has taken different views about this matter at different times and in different places. If, as we should, we recognize the rule against the use of involuntary statements as part of a constellation of rules protecting the accused's free choice whether to cooperate or not with state authorities in the investigation of the crime alleged against him or her, then it is very hard to talk about the voluntariness of a statement without considering those other rules and how they affected the accused's choice (for example, the role that the right to counsel might play in protecting the right against self-incrimination). Although an improvement in accuracy is not a sufficient justification for abolishing the right against self-incrimination, the likely effect of these rules on the accuracy of fact-finding has a role to play in defining the relevant legal standards for what counts as self-incrimination. Within some margin for debate about what makes a choice free, if a court's best guess is that a given way of construing the right would improve accuracy, then that would count in favor of construing it that way. This kind of justification could not be used to construe the rule so as to admit statements made under torture, nor to exclude unprompted confessions made by people who are not under any pressure from the investigative authorities. Nor could it be used (as, unfortunately, it sometimes has been) as a kind of general justification for a more generous rule of admissibility. But it might have some role to play in, for example, a court's determination of the way various interrogation techniques affect voluntariness.
Other procedural rights can also be specified in this way, that is, with a view to improving accuracy. Consider again the right to finality discussed above. As we saw, in the United States, the right against double jeopardy is interpreted as preventing the prosecution from appealing an acquittal, even if the trial was legally flawed. Laudan argues at length that this interpretation of the right reduces the fact-finding accuracy of the criminal process.Footnote 63 My guess is that he is right: permitting the prosecution to appeal acquittals tainted by legal error would probably reduce the number of incorrect acquittals without substantially increasing the number of wrongful convictions. But the remedy for this problem is not to abandon the right to finality, to permit the prosecution to retry acquitted persons willy-nilly; it is to try to construe the right to finality so as to permit this kind of improvement in accuracy, for example by permitting the prosecution to appeal on the ground of legal error alone and requiring the prosecution to show that if there was a legal error, it would likely have affected the verdict.
So, if there is a right to accuracy in fact-finding, it is not the same kind of right as the right to be presumed innocent or the right to finality. No one can have a right to any particular level of fact-finding accuracy in their own case; they can only have a right to the conscientious application of whatever rules are in place to govern fact-finding. Nevertheless, accuracy in fact-finding is an appropriate goal for the legislatureFootnote 64 to pursue because of its contribution to the rule of law. And so inaccuracy in fact-finding is an appropriate subject for complaint, and the relative weights to be (notionally) assigned to the various terms of the procedural loss function are appropriate matters for policy debate. Thus, if it is necessary to speak of a right to accuracy in fact-finding, it would be the same kind of right as the right to a fair system of taxation. When I complain of a particular rule of fact-finding, I cannot complain merely of its effects on me but must take into account its effects on the system of fact-finding as a whole, just as I can't object to a change in a tax rule on the basis that it causes me to pay more tax or my neighbor to pay less, but on the basis of its economic and fiscal effects in general. Nevertheless, while there is plenty of room for legislative choice about rules of evidence and procedure, those choices are subordinate to whatever procedural rights people do have, such as the right to be presumed innocent, the right to finality, and the right to a fair hearing.
V. CONCLUSION
People have procedural rights for the same reason that they have substantive rights. Each person is a free, purposive, and choosing agent who pursues their purposes constrained by legal rules and principles that ensure mutual consistency in everyone's pursuit of their purposes. Those rules and principles define a set of private law and public law rights that is suitable for such purposes. Disputes about the application of the law, and about the facts on which application of the law depends, are inevitable. The process for resolving such disputes must be consistent with the freedom of the legal persons whose disputes are being resolved, and it must resolve those disputes with a reasonable degree of factual accuracy; if it does not do those things, it is not a process that is suitable to the legal order. The requirement of consistency with legal personhood gives rise to the two procedural rights that I have considered in his paper: the right to procedural fairness and the right to finality. There are others, which I have not discussed here. The requirement of a reasonable degree of accuracy means that, within the constraints imposed by procedural rights, the rules of evidence and procedure should be designed with a view to reducing the normative losses generated by inaccuracy in fact-finding. Although it is difficult, perhaps impossible, to calculate the normative losses created by any given set of procedures or to anticipate the impact of a change in procedures, the idea of normative losses can nevertheless serve as a heuristic device for assessing different ways of specifying the rules of evidence and procedure that are otherwise consistent with the procedural rights of individuals.
