(Originally published in the Cornell Law Review, Volume 88, Number 6, July 2003. Republished with the author's permission.)

Criminal procedure has for too long treated itself as a subset of constitutional law¹, in the process distancing itself from substantive criminal law. While substantive criminal law sometimes discusses how well rules deter, rehabilitate, or exact retribution, these substantive values are largely absent from criminal procedure. Instead, criminal procedure seems to care only about whether procedures are efficient, constitutional, fair, and accurate at finding the truth. This artificial separation is unfortunate. Criminal procedure is not simply a subset of constitutional law. It is a sibling of criminal law, though our narrow curricular blinders keep us from seeing the import of this fact. A procedure may be constitutional, efficient, procedurally fair, and even accurate but deeply unwise. If the procedure undermines important values of the substantive criminal law, we should reject it no matter how efficient it is.

The divorce of procedure from substance manifests itself, for example, in guilty plea procedures. (I use the terms substance and substantive criminal law to distinguish the body of law that defines crimes from the procedures used to enforce them. My focus is not on the actus reus and mens rea elements of particular crimes, but rather on the justifications for punishment that underlie these crimes.) The standard defense of plea bargaining is that guilty pleas save time and money, reduce uncertainty, and empower parties by promoting freedom of choice.² These values are procedural values, focusing on the choices and costs parties face in court. The standard critique of plea bargaining is that guilty pleas undercut proof beyond a reasonable doubt, adversary hearings, and other procedural safeguards. Once again, most of these objections rest on process values rather than the values of the substantive criminal law. In this article, I do not wade into the broader debate over the desirability of plea bargaining. Instead, I take as a given that guilty pleas and plea bargaining will persist for the foreseeable future. I use the plea-bargaining literature to illustrate the foci of the proceduralists on both sides: efficiency and autonomous choice versus accuracy and fairness.

This article challenges the proceduralist approach to criminal procedure, using two subsets of pleas as case studies. First, the law has long allowed defendants to plead nolo contendere. This means that they refuse to admit guilt but accept punishment as if guilty. More recently, the Supreme Court has approved so-called Alford pleas, in which defendants plead guilty while simultaneously protesting their innocence.³ Far from criticizing these practices, Judge Frank Easterbrook and most other scholars praise these pleas as efficient, constitutional means of resolving cases.⁴ Even Albert Alschuler, a leading critic of plea bargaining generally, supports Alford pleas. He views them as a lesser evil, a way to empower defendants within a flawed system. As long as we have plea bargaining, he maintains, innocent defendants should be free to use these pleas to enter advantageous plea bargains without lying. And guilty defendants who are in denial should be empowered to use these pleas instead of being forced to stand trial.⁵ Once again, the terms of the debate are proceduralist: efficiency and autonomy versus accuracy and fairness.

I dispute this conventional wisdom. Alford and nolo contendere pleas, I contend, are unwise and should be abolished. These procedures may be constitutional and efficient, but they undermine key values served by admissions of guilt in open court. They undermine the procedural values of accuracy and public confidence in accuracy and fairness, by convicting innocent defendants and creating the perception that innocent defendants are being pressured into pleading guilty. More basically, they allow guilty defendants to avoid accepting responsibility for their wrongs. Guilty defendants' refusals to admit guilt impede their repentance, education, and reform, as well as healing of victims. In addition, pleas without confessions muddy the criminal law's moral message. Both kinds of pleas, but especially Alford pleas, equivocate: one might call them guilty-but-not-guilty pleas. ⁶ They permit equivocation and ambiguity where clarity is essential. This equivocation undermines denunciation of the defendant and vindication of the victim and the community's moral norms. Sacrificing these substantive goals of the criminal law is too high a price for an efficient plea procedure. (There is little point in having procedures that undercut substance, as the point of procedure is to serve substance. Yet substantive values are, for the most part, not even on the proceduralists' radar screens.)⁷ Thus, guilty pleas should be reserved for those who confess.⁸ Jury trials should serve not only to acquit innocent defendants, but also as morality plays to teach guilty defendants and vindicate their victims and the community's moral norms. The criminal law's norms include honesty and responsibility for one's actions, so criminal procedure should not let guilty defendants dishonestly dodge responsibility and the truth.

Consider this prominent example, which illustrates why unequivocal guilty-plea confessions serve these values better than equivocal Alford and nolo pleas: In the 1970's, Kathleen Soliah belonged to the Symbionese Liberation Army, a radical San Francisco group that kidnapped Patricia Hearst and tried to kill government officials. Soliah fled to Minnesota and changed her name to Sara Jane Olson. For years, she denied belonging to the Symbionese Liberation Army or taking part in trying to bomb two police cars in 1975. Her lawyer expressed interest in negotiating an Alford or nolo contendere plea, but the judge and prosecutors said they would not countenance such a plea.⁹ Here, Olson's judge exercised his discretion in refusing to accept an Alford or nolo contendere plea, but judges elsewhere often allow such pleas. See infra Part II.C. If a different judge had had the case and allowed an Alford plea, the outcome would have been very different, as the text goes on to explain. So, on October 31, 2001, Olson clearly and unequivocally pleaded guilty to attempting to bomb the two cars. Immediately afterwards, however, Olson told reporters that she had pleaded guilty to crimes of which she was innocent. Prosecutors speculated that Olson had changed her story to please her friends and family who had maintained her innocence.¹⁰

Olson's judge refused to countenance this express and instantaneous contradiction, noting that "the integrity of the criminal justice system is at stake."¹¹ He called Olson in for another hearing and asked her whether she wanted her plea to stand. At that hearing, the judge confronted Olson and asked her, clearly and explicitly, if she was in fact guilty. She twice said yes and reaffirmed her plea.¹² Eight days later, Olson again publicly disputed her guilt and moved to withdraw her plea.¹³ At the next court hearing, the judge noted that Olson found it psychologically very difficult to admit her crime to herself, her family, and her supporters. Relying on her previous admissions and pleas of guilt, the judge denied Olson's motion to withdraw her plea. ¹⁴ Only after this final ruling did Olson tremble with emotion and say she was sorry for harming others.¹⁵

An Alford or nolo plea in this case would have undercut important procedural and substantive values and norms. If Olson had entered an Alford plea and never admitted guilt, it would have been wrong to punish her despite her protestations of innocence without an authoritative trial verdict. Instead of eventually apologizing, she might well have persisted in her denials to herself and to others. Continued denials would have led her friends and family and the public to doubt the justice of the system. Punishment in these circumstances would have undercut the norm of punishing only those known to be blameworthy. In addition, consistent protestations of innocence would have hindered closure for victims and the community. Here, in contrast, Olson made clear admissions of guilt in court, so her later denials were less credible. The public could more easily believe that she had falsely protested her innocence to save face. The court was able to justify its ruling by pointing to Olson's earlier admissions in open court on the advice of counsel. Here, the court's action vindicated the norm of not going back on one's word. And, after the judge confronted her with her earlier admissions in court, Olson took the first steps toward apology and reconciliation. In short, Olson' admissions of guilt in open court were much firmer bases for conviction, repentance, and closure than an Alford or nolo plea would have been.

The remainder of this article consists of four parts. Part I summarizes the academic debate over plea bargaining, showing how it embodies criminal procedure's emphasis on procedural values. Part II reviews the doctrines that allow Alford and nolo contendere pleas and the scholarly articles supporting these doctrines, most of which support these pleas on proceduralist grounds. It also discusses how, when, and why lawyers and clients use these pleas to avoid admissions of guilt. In Part III, I object that these procedures risk convicting innocent defendants and create the perception that innocent defendants are being convicted. The analysis rests on the conventional procedural values of accuracy and perceived accuracy, but it looks at them through a moral lens. This kind of moral argument is almost unheard of in proceduralist literature, a clue that procedure is adrift from the moral moorings of the substantive criminal law.

Part IV moves beyond Part III's conventional procedural values to substantive-criminal-law values. In particular, it critiques Alford and nolo contendere pleas from a moral, didactic perspective. Even if these pleas were perfectly accurate and were so viewed, they would undercut reform, moral education, and vindication of victims and the community's moral norms. Many guilty defendants are in denial and find it hard to admit their crimes to others or even to themselves. For them, Alford and nolo contendere pleas are easy ways to remain in denial and avoid the painful processes of confession or trial. Trials, though less efficient and swift, are better at breaking through these denials and beginning the process of reform and healing. And regardless of how defendants respond, convictions after trial vindicate victims, express outrage, and drive home the wrongfulness of crimes to defendants. In other words, the social meaning of a jury verdict or guilty plea is much stronger and clearer than an Alford or nolo plea's muddy message. Thus, legislatures should abolish Alford and nolo contendere pleas. Until they do so, prosecutors should oppose them and judges should exercise their discretion to reject them. This article concludes with thoughts on restructuring plea procedures and lawyers' and judges' roles to serve the norms and values of the substantive criminal law.

I. THE PROCEDURALIST APPROACH TO PLEAS

The dominant approach to guilty pleas and plea bargaining focuses on procedural values such as speed, cost, efficiency, autonomy, accuracy, and certainty. This proceduralist focus largely ignores substantive values such as reform, education, retribution, and vindication of victims and social norms. My purpose here is not to take on all pleas and bargains; I take these as given. Rather, my aim is to illustrate how proceduralists' emphasis on procedural values comes at the expense of substantive-criminal-law values. As Part II explains, this proceduralist approach to pleas and bargains in general carries over to Alford and nolo contendere pleas in particular.

The Supreme Court, for example, has endorsed guilty pleas because they save time and money and confer advantages upon both prosecutors and defendants.¹⁶ The Court's sole concern is to ensure procedural safeguards such as adequate counsel, knowing and voluntary waivers of rights, and sufficient factual bases.¹⁷ Similarly, the criminal bench and bar like plea bargaining because it saves time and money, caps defendants' sentences, expedites inevitable convictions, and disposes of large caseloads.¹⁸ Of course, this generalization does not hold true of every single lawyer. A few take criminal court appointments precisely because they want trial experience. Others may prefer to work more hours so that they can claim larger fees (when fees are computed on an hourly basis). See United States v. Diaz, 802 F. Supp. 304, 312 (C.D. Cal. 1992) (complaining about this practice of "fee churning"). These approaches, however, are not the norm. Though pleas may incidentally serve substantive values (such as quicker incapacitation or rehabilitation), the emphasis is on saving time and money and allowing the parties to choose.

Scholars who support plea bargaining likewise do so because pleas promote procedural values such as speed, cost, efficiency, and free choice. Judge Frank Easterbrook, one leading proponent, views criminal justice as a market system that lets parties sell procedural rights in exchange for advantageous concessions.¹⁹ According to Easterbrook, Robert Scott, and William Stuntz, these sales of rights promote autonomy and efficiency, reduce uncertainty, and save time and money.²⁰ Scott and Stuntz's concerns are primarily procedural ones: they recognize the need for special procedural safeguards to prevent duress, mistake, unconscionable pressures, and uninformed decisions.²¹ In other words, they view the role of plea procedures as giving defendants the information and freedom they need to further their own interests and desires. One might call this the autonomy model: it takes defendants' current desires as a given instead of seeking to reshape or trump those desires.

Critics of plea bargaining focus on classic procedural values such as accuracy and procedural fairness. Albert Alschuler's and Stephen Schulhofer's objections are numerous, but most fall into two categories: First, plea bargaining undermines structural safeguards by letting prosecutors usurp the neutral judicial role, letting defense counsel cut corners, and avoiding public trials.²² One might call this the institutional or adversarial perspective. Second, bargaining undercuts accuracy, equal treatment, fairness, and perceptions of fairness, by subverting proof beyond a reasonable doubt and other rights, putting innocent defendants at risk.²³ One might call this the defendants'-rights perspective.

To be sure, a few cases and commentators have suggested in passing that plea bargaining might serve or hinder values of the substantive criminal law. With the notable exception of one Alschuler article, none of these discussions of substantive values occupies more than a few pages.²⁴ All in all, these sporadic references to deterrence, incapacitation, retribution, and rehabilitation are peripheral to the academic and judicial debate.

In short, the recent plea-bargaining debate illustrates the proceduralist methodology that pervades criminal procedure. Most recent considerations of plea bargaining stand or fall on procedural values: the autonomy, accuracy, efficiency, fairness, and perceived fairness of the process.²⁵ They pay little heed to rehabilitation, reform, education, and other substantive-criminal-law values. As Part II explains, this proceduralist approach to pleas in general carries over to Alford and nolo contendere pleas in particular.

II. THE STATUS QUO ON ALFORD AND NOLO

The proceduralist approach to pleas and bargains pervades discussions of Alford and nolo contendere pleas. Recall that these pleas are the functional equivalent of guilty pleas, except that defendants do not admit guilt and (in nolo pleas) are not estopped by their pleas in later litigation. Most courts and commentators support these pleas, stressing their efficiency and speed and the desirability of letting defendants choose to protect their privacy and dignity. The few critics who discuss these pleas in passing generally emphasize the danger that innocent defendants may falsely plead guilty. This emphasis on choice, efficiency, and accuracy exemplifies the classic proceduralist justifications for plea bargains discussed in Part I. Section A summarizes the law governing these types of pleas. Section B surveys the generally favorable academic commentary on these pleas. Section C looks at how often defendants use these pleas, in what kinds of cases, and why. It also contrasts academics' generally favorable reaction to these pleas with notes of skepticism expressed by judges and prosecutors about these pleas.

A. The Law of Nolo Contendere and Alford Pleas

At common law, a defendant could ask the court to impose a merciful sentence without confessing guilt and without estopping himself from later pleading not guilty on the same facts.²⁶ In modern times, this became the formal plea of nolo contendere, which admits guilt for purposes of the present case but creates no estoppel.²⁷ Today, the Federal Rules of Criminal Procedure allow defendants to plead nolo contendere with the permission of the court.²⁸ Most states likewise allow nolo contendere pleas (sometimes called no contest), though in many states these pleas require the court's consent.²⁹

Defendants now have another way to plead guilty without admitting guilt: the Alford plea. Henry Alford was charged with first-degree murder, a capital crime, and faced strong evidence of guilt. Rather than going to trial, he pleaded guilty to second-degree murder, a non-capital crime, while protesting his innocence.³⁰ In North Carolina v. Alford, the Supreme Court held that defendants may knowingly and voluntarily plead guilty even while protesting their innocence. The Court noted that judges must find a sufficient factual basis before accepting these pleas, but the State provided one here. It put on two incriminating witnesses, who testified that Alford had left his house with a gun saying he would kill the victim and had returned saying he had killed the victim. The Court also noted that Alford's plea was like a plea of nolo contendere. If defendants can plead nolo contendere while refusing to admit guilt, the Court saw no reason to bar Alford from pleading guilty while protesting his innocence. The Court also suggested that Alford's decision to plead was a reasonable choice to cap his maximum sentence and that courts should honor this choice. While these pleas are not forbidden by the Constitution, neither are they required. Because defendants have no right to plead guilty, judges may refuse to accept Alford pleas, or states may forbid them by statute or rule.³¹ Most states have followed suit and permitted Alford pleas (sometimes called best-interests pleas).³²

Note that there are two main differences between Alford and nolo contendere pleas: First, nolo contendere pleas avoid estoppel in later civil litigation, while Alford pleas do not. Second, defendants who plead nolo contendere refuse to admit guilt, while Alford pleas involve affirmative protestations of innocence. By and large, however, Alford is simply a new extension of the age-old nolo plea.³³ The timing of this expansion of the law three decades ago may be no coincidence. Alford fit well with the modern liberal emphasis on freedom of contract, autonomy, and informed choice.

B. The Scholarly Literature

Those commentators who have considered Alford and nolo contendere pleas have endorsed them for varying reasons. Probably the most common argument in favor of them is that they are a tool for resolving cases efficiently and cheaply.³⁴ Judge Frank Easterbrook and others support these pleas because they further the interests of defendants (even innocent defendants) who want to avoid worse outcomes at trial.³⁵ In other words, these pleas promote autonomy: they give defendants another, easier choice, one that may benefit their interests. Others state that these pleas protect defendants' dignity, privacy, and autonomy by obviating humiliating public admissions of guilt.³⁶ Another argument is that nolo pleas protect "the respectable citizen" who is "technically guilty" but does not deserve civil disabilities (such as losing the rights to vote and hold office).³⁷ And some commentators claim that Alford pleas foster openness between lawyer and client. If there were no Alford pleas, they claim, innocent defendants would lie about their guilt to their lawyers in order to reap the benefits of pleading guilty.³⁸

Even one prominent critic of plea bargaining endorses Alford pleas. Albert Alschuler, as mentioned above, would prefer to abolish plea bargaining entirely. But given that plea bargaining exists, Alschuler reluctantly endorses Alford pleas.³⁹ Many defendants are unwilling to admit guilt because "`psychological obstacles,'" egos, and shame get in the way.⁴⁰ Yet it is in their interests to plead guilty. For these defendants, the Alford plea is a necessary psychological "crutch."⁴¹ Even innocent defendants, he argues, should be able to choose Alford pleas if they decide that pleading is in their best interests.⁴² He notes that even after Alford, many defense lawyers and judges refuse to permit Alford pleas.⁴³ He claims that if lawyers and judges insist on admissions of guilt, defendants will lie to their lawyers and the court, and defense counsel will pressure clients to confess or lie. Though he finds Alford pleas distasteful and offensive, he nonetheless supports them as more honest and fair and less hypocritical. They keep defendants from having to lie, keep defense lawyers from coercing confessions, and avoid forcing defendants into disadvantageous trials.⁴⁴

The few critics of Alford pleas rest their criticisms primarily on proceduralist grounds. John Langbein and others argue that Alford pleas undercut proof beyond a reasonable doubt and risk allowing innocent defendants to plead guilty.⁴⁵ Another proceduralist objection is that Alford pleas risk being involuntary, as reluctant defendants are likely caving in to coercive pressures to plead.⁴

There is, however, a different criticism in the writings of David Wexler and Bruce Winick. Wexler and Winick suggest that Alford and nolo pleas to sex offenses are undesirable because they allow offenders to remain in denial. Instead, judges should refuse to accept these pleas, which would in turn force defense lawyers to confront their clients with the facts and break down their illusions and denials. Wexler and Winick focus on rehabilitation of offenders, but they do not discuss other substantive values nor the effect of these pleas on victims or the public.⁴⁷

C. Who Uses These Pleas, When, and Why?

The only statistics on nolo contendere pleas come from federal cases. In the year ending September 30, 2000, 0.5% of all federal defendants pleaded nolo contendere.⁴⁸ These pleas tended to cluster in certain categories of cases. For example, 6% of those charged with drunk driving and other traffic offenses pleaded nolo contendere.⁴⁹ Nolo contendere pleas were also more likely to be used for white-collar crimes such as fraud, counterfeiting, food and drug, and environmental laws.⁵⁰ Nolo contendere pleas were used in about 5% of federal sex offenses.⁵¹ Anecdotal evidence suggests that nolo contendere pleas are popular in antitrust cases.⁵²

Of course, federal numbers risk being unrepresentative because most cases and types of crimes are handled primarily at the state level. Because no state statistics are available, I conducted a series of Westlaw searches for cases involving nolo contendere pleas. This survey revealed that nolo contendere pleas are even more common than these federal numbers suggest. This survey found more than 18,500 cases involving nolo contendere pleas. Of the relevant search results in a random sample, 26% involved drug crimes, 25% involved property crimes (including embezzlements), 23% involved violent crimes, 21% involved sex crimes, and 14% involved white-collar crimes.⁵³ Note that the figures in the text add up to more than 100% because some cases involved nolo contendere pleas to multiple types of charges.

Since the Court decided Alford, Alford pleas have become quite common, especially in sex-offense cases. There are no statistics on Alford pleas, so I conducted a series of Westlaw searches. These searches found almost 2500 cases that involved Alford pleas. Of the relevant search results in a random sample, 27% of Alford pleas involved sex offenses, 27% involved other violent offenses, and 12% involved white-collar offenses.⁵⁴

To get a sense of when and why lawyers and defendants use these pleas, I interviewed thirty-four veteran prosecutors, public and private defense lawyers, and judges. I used Westlaw searches to pick states that appear to use Alford or nolo pleas frequently (Louisiana, Michigan, Missouri, Pennsylvania, and Ohio). I contrasted these states with states that forbid both kinds of pleas (Indiana and New Jersey). I followed no scientific method, and of course my sample size was far too small to generate statistically significant results. My methodology was journalistic and impressionistic; it replicated on a much smaller scale the surveys on which Albert Alschuler built his famous articles on plea bargaining.⁵⁵

According to the defense lawyers interviewed, many if not most defendants are initially reluctant to admit guilt. Defense counsel work with defendants, confront them with the evidence, and bring most around to where they will admit guilt. A small minority of clients remain unwilling to admit guilt even when it would be in their interests to do so. Some go to trial, but others enter Alford or nolo pleas. Lawyers estimated that a few percent of cases were resolved by one of these pleas.

When I asked defense counsel, prosecutors, and judges why they thought these defendants would not admit guilt, their answers tended to converge. The number one barrier to a classic guilty plea is fear of embarrassment and shame before a defendant's family, friends, and others. Many of these defendants can admit guilt to their lawyers but are too ashamed to do so in front of their family. Defense lawyers work hard to convince families that their children should plead, and do all they can to reduce the shame of pleading. One defense lawyer even schedules pleas for late Friday afternoons and misleads clients' families about plea dates, so that his clients can plead more easily in empty courtrooms. Even these shame-reducing measures, however, are not enough for some defendants.

After shame, the reason cited most frequently is psychological denial. Some defendants are in denial and refuse to admit guilt to themselves. Some lawyers mentioned collateral consequences: An admission of guilt may hurt in a later child-custody battle. Admissions may also scare off prospective employers, whereas Alford and nolo contendere pleas make it easier to reassure employers by denying guilt. And nolo pleas avert estoppel in collateral civil litigation (especially in automobile accidents). Also, defendants may have been intoxicated and unable to remember the facts. Several lawyers opined, however, that most claims of lost memory or fear of collateral consequences are fig leaves to justify these pleas.⁵⁶ The real, hidden reason is more often feelings of shame or guilt. Almost all interviewees agreed that completely innocent defendants use these pleas infrequently: their descriptions ranged from "occasionally" to "extremely uncommon" to "[in]significant" to "very rare."⁵⁷ For example, one long-time public defender estimated that he had seen no more than five to ten innocent defendants use these pleas over the last sixteen years. A few defense lawyers did suggest that some defendants enter Alford or nolo pleas to crimes more serious than the ones they committed.

Opinions differed on the kinds of cases in which defendants use these pleas. A few interviewees said that they could not generalize, as they use Alford and nolo contendere pleas in a variety of cases. But most thought certain kinds of cases were most likely to involve Alford and nolo pleas. By far the largest category of cases is sex offenses. Sex offenders are often in denial and fear shame, rejection by families and girlfriends, and violence by other prisoners. ⁵⁸ A second category is crimes against children or the elderly, especially sex crimes. These include child molestation, incest, and rape, but also non-sexual child abuse and neglect. Others mentioned heinous murders, domestic assaults, batteries, crimes of dishonesty, drunk driving, drugs, and auto accidents (to avoid estoppel in tort suits).

Every defense lawyer whom I interviewed approved of these pleas. They use them infrequently, as a last resort, a tool for difficult defendants who simply will not admit guilt. Defense lawyers reported that most but not all prosecutors are amenable to Alford and nolo pleas, but that judges vary widely and many will not accept them. (They reported that Alford and nolo pleas require judges' consent, as well as prosecutors' consent if the pleas are part of plea bargains.) In other words, prosecutors and judges are more ambivalent. On the one hand, many see these pleas as efficient ways to dispose of cases and reduce staggering dockets. Reducing dockets is, for many, the top priority. On the other hand, they fear that pleas without admissions of guilt are more vulnerable to appeal or collateral attack, which undercuts finality. Some prosecutors and especially judges dislike the message that these pleas send. Some judges view criminal justice as a morality play in which defendants should confess and apologize, so that victims see justice done and begin to heal. Pleas without confessions, on this view, leave victims frustrated and defendants defiant and resistant to treatment. Two defense lawyers suggested that victims or their families sometimes press prosecutors to oppose Alford pleas, because they want admissions of guilt and apologies. And some prosecutors and judges worry that pleas by defendants who deny or equivocate undermine public confidence. Defendants can deny guilt out of court (the Sara Jane Olson maneuver), leading family, friends, and the public to suspect injustice. As one jurist put it, unequivocal pleas and trials "suppor[t] public trust in the institutions" by impeding later denials.⁵⁹ In short, prosecutors and judges sometimes oppose Alford and nolo pleas on both consequentialist and other moral grounds.

What happens when the law forbids Alford and nolo pleas, or judges refuse to allow them, or prosecutors refuse to enter plea bargains without admissions of guilt? Though some of these cases go to trial, many defendants come around and eventually admit guilt (as Sara Jane Olson did). Judges and counsel in states that forbid these pleas agreed on this point: A majority if not most defendants who deny guilt at plea hearings, when required to admit guilt or go to trial, in the end admit guilt.⁶⁰ Statistics, though not definitive, suggest that rejection of these pleas does not create a huge number of trials. A National Center for State Courts report collected criminal case-processing statistics for 22 states in the year 2000. It found that on average 3.3% of cases were resolved at trial, 57.2% were resolved by pleas, 23.3% were resolved by dismissals or nolle prosequis, and 16.2% had other unspecified dispositions. The two jurisdictions that forbid Alford and nolo contendere pleas had comparable figures, with trial rates only slightly higher than average. In New Jersey, 3.9% of criminal cases went to trial, 68.4% pleaded, 15.2% were dismissed or nolled, and 12.4% had other dispositions. In Indiana, 4.3% of criminal cases went to trial, 57.3% pleaded, 35.5% were dismissed or nolled, and 3.0% had other dispositions. Nat'l Center for State Courts, Examining the Work of State Courts, 2001, at 63 (2001). One judge estimated that while he has balked at no-contest pleas for ten years, only one or two defendants have gone to trial rather than admit guilt. For example, when a judge refused to let two prostitution defendants plead nolo contendere, they went down the hall, came back a few minutes later, and pleaded guilty. In another case, when a prostitute's customer refused to admit guilt and the judge refused to accept his guilty plea, the defense lawyer conferred with his client and he later admitted guilt.

Other evidence corroborates my finding that prosecutors and judges are deeply ambivalent about Alford and nolo contendere pleas. For example, even though federal law permits Alford pleas, U.S. Department of Justice policy discourages them. Because the public may not understand how a defendant who claims innocence can plead guilty, the public may suspect prosecutorial overreaching. Thus, federal prosecutors may not agree to Alford pleas absent approval by the main Department of Justice in Washington. If defendants try to enter Alford pleas without prosecutorial approval to less than all counts, federal prosecutors must discourage them by refusing to dismiss the remaining counts.⁶¹ Indeed, the Antitrust and Tax Divisions go further and oppose all Alford and nolo contendere pleas. The Tax Division's explanation is that such pleas undercut collateral estoppel and mislead the public into thinking that the Government's case is weak.⁶² Perhaps these central policies are high-level efforts to send public messages and enforce consistency by riding herd on line prosecutors. Line prosecutors are tempted to lighten their own workloads by using every tool available to secure pleas, because the long-term impact of any one plea is slight. In contrast, high-level policymakers are more likely to consider systemic, long-term ramifications. Other prosecutors likewise disfavor Alford and nolo pleas.⁶³ Defendants may nonetheless be able to enter Alford and nolo pleas over the Government's objections.⁶⁴

Several state courts have followed suit and forbidden Alford pleas. The Supreme Court of Indiana has held that judges may not accept guilty pleas accompanied by protestations of innocence.⁶⁵ The court suggested that Alford pleas risk being unintelligent, involuntary, and inaccurate.⁶⁶ Another reason for the Indiana rule is that Alford pleas undercut public respect for the justice system.⁶⁷ Michigan and New Jersey courts agree and also forbid Alford pleas.⁶⁸ Arizona permits Alford pleas but disfavors them, for fear that innocent defendants will plead guilty or that the public will lose confidence in the justice system.⁶⁹ Individual judges in other states disfavor Alford pleas as well.⁷⁰ And Alschuler found (contrary to my finding) that many defense lawyers refuse to allow clients to plead guilty if they claim innocence.⁷¹

This reluctance to accept speedy Alford and nolo contendere pleas suggests that they are deeply troubling. Though these pleas are efficient, they disregard other important values--both procedural values (such as accuracy) and substantive values (such as reform, education, and expressive condemnation). Parts III and IV develop these arguments further.

III. ACCURACY AND PERCEIVED ACCURACY

Efficiency is a value in criminal procedure, but it is not the only nor even the most important value. More important is the accuracy of the system's results, and in particular its accuracy in freeing innocent defendants. Yet while our system goes to great lengths to protect innocent defendants at trial, it perversely makes it too easy for them to plead guilty by allowing Alford and nolo contendere pleas. Innocent defendants whose scruples might otherwise prevent them from pleading guilty can use these pleas to do so more easily. Even if innocent defendants want to plead guilty, the law should not go out of its way to promote these unjust results.

This Part argues that Alford and nolo pleas disserve the conventional procedural values of accuracy and perceived accuracy. Section A contends that allowing innocent defendants to plead is wrong and rebuts the contrary arguments of Alschuler, Easterbrook, and others. Innocent defendants who plead are often overestimating their likelihood of conviction at trial, so the law should encourage them to persevere and win acquittals. Moreover, Easterbrook's utilitarianism ignores the moral imperative to avoid knowingly facilitating injustice. (This moral reasoning goes beyond the narrow utilitarianism found in most proceduralist literature.) Section B considers how Alford and nolo contendere pleas undermine public perceptions of the justice system's accuracy and fairness. Part IV then goes beyond Part III's proceduralist approach, explaining how these pleas violate important values of the substantive criminal law.

A. Convicting Innocent Defendants Is Wrong

It should go without saying that convicting innocent defendants is wrong. Thus, the law should hinder these convictions instead of facilitating them through Alford and nolo contendere pleas. Nonetheless, Easterbrook, Alschuler, and others favor Alford and nolo pleas in part because they enable innocent defendants to plead guilty without lying.⁷² Some of these commentators, notably Easterbrook, assume that increasing the range and ease of choices is always good. Increasing the ease of convicting innocent defendants is a vice, however, not a virtue. Innocent defendants who would otherwise balk at admitting guilt favor Alford and nolo contendere pleas as easier ways out. The law should not afford these easier ways out, in the hopes that innocent defendants will balk at admitting guilt and persevere to gain acquittals or dismissals.⁷³ Even if innocent defendants think that pleading guilty is in their interests, the law should make it harder for them to do so. Doing so would minimize both actual and perceived injustices.

Easterbrook responds that innocent defendants will plead guilty only when the expected sentence at trial, discounted by the probability of acquittal, is greater than the plea terms offered. Defendants who are advised by competent counsel and usually have private knowledge of the facts will be in a good position to assess their own chances at trial. Trials are imperfect and sometimes convict innocent defendants. So, the argument goes, innocent defendants benefit by having the option of pleading guilty whenever they might be convicted at trial.⁷⁴ As Alschuler puts it, "both courts and defense attorneys should recognize a `right' of the innocent to plead guilty. So long as a defendant has something to gain by entering a plea agreement, it is unfair to deny him the choice."⁷⁵

Easterbrook's line of argument mistakenly treats innocent defendants as fully informed, autonomous rational actors. Many defendants receive poor advice from overburdened appointed counsel of varying quality whose caseloads and incentives lead them to press clients to plead guilty.⁷⁶ Poor counsel would also hinder these same defendants at trial. But the financial incentives to encourage pleas would not affect trials as much because of the large fixed costs involved, and lawyers' desire to preserve their reputations by prevailing at trial would counteract financial incentives to cut corners. Schulhofer, Criminal Justice Discretion as a Regulatory System, supra, at 56-59. In addition, criminal discovery is not nearly as extensive as civil discovery, hampering defendants' accurate assessments of their prospects at trial.⁷⁷ And innocent defendants may plead guilty because of pressure or misinformation. As a result, their pleas may not be fully intelligent and voluntary.⁷⁸ Thus, innocent defendants who want to enter Alford or nolo pleas are likely overestimating their risk of conviction at trial.⁷⁹ Those poor enough to qualify for overburdened appointed counsel and those of low intelligence are most likely to make these mistakes. The result may well be troubling disparities based on wealth, mental capacity, and education. The law should instead encourage these innocent defendants to persevere and win acquittals.⁸⁰ So a paternalistic concern for poorly educated and misled innocent defendants might justify abolishing or severely restricting Alford and nolo pleas.

There is also a deeper moral objection to Easterbrook's purely utilitarian argument. One's moral intuition should recoil at the thought of allowing, indeed facilitating, the conviction of innocent defendants. It is all the more troubling to trumpet this fact as an advantage of Alford and nolo pleas. Not all of ethics is reducible to a consequentialist calculus. There is something deeply troubling about knowingly facilitating injustice, something worse than allowing it to happen inadvertently. No promise of good consequences can erase the repugnance of promoting an evil in the hope of averting a worse evil. To use Dostoyevsky's example, no hope of good consequences can justify society's murdering a single innocent child. ⁸¹ Kant would agree that society cannot knowingly facilitate the punishment of those who do not deserve it, even if they agree to it.⁸² Second, society must not consciously promote guilty and nolo contendere pleas by innocent defendants. Unless one "has committed a crime" and been found "guilty and punishable," no amount of benefits can justify punishment. Immanuel Kant, The Philosophy of Law *195-97 (W. Hastie trans. 1887) (corresponding to pages *331-*332 in The Metaphysics of Morals). Even though the justice system may accidentally punish some innocent defendants, it has no right to consciously facilitate this injustice.

It is doubtful that our criminal justice system charges and prosecutes large numbers of innocent defendants. There are doubtless some, however,⁸³ and they may be tempted to use Alford and nolo pleas instead of going to trial. Though it is impossible to know how many innocent defendants are at stake, many of the lawyers whom I interviewed thought that innocent defendants occasionally used these pleas. And the law should take pains to reduce the chances of convicting even these few innocent defendants.

One can rarely prove directly that a particular defendant who pleaded guilty or nolo contendere was in fact innocent. But there is troubling anecdotal evidence that innocent defendants do use these pleas. For example, in the notorious Wenatchee case, twelve defendants entered Alford or nolo contendere pleas to child molestation-related charges. Two of them later adduced evidence that a complaining child had never been abused but had falsely incriminated them because of coercive police interrogation. After the child recanted, these two defendants were eventually allowed to withdraw their pleas.⁸⁴ Another case in which an innocent man entered an Alford plea occurred in Virginia. Abusive interrogation pressured a developmentally disabled man, David Vasquez, to confess and enter an Alford plea to a murder. Investigators later concluded that the crime fit the modus operandi and pattern of another man, a serial killer, and persuaded the Governor to pardon Vasquez. Brooke A. Masters, Lucky Release from a Life Behind Bars: Va. Man Served 5 Years-Under Plea Agreement-Before Real Murderer Was Found, Wash. Post, Apr. 28, 2000, at A23. Extensive media coverage later suggested that the police investigator had led children to fabricate these and thousands of other allegations of sexual abuse.⁸⁵

One cannot know how many similar cases, perhaps less sensational or on a smaller scale, never come to light, veiled by defendants' waivers of appellate rights. Even if only a few innocent defendants enter Alford or nolo pleas, these injustices are grave and deeply troubling.

B. Public Perceptions of the Justice System

The justice system cannot consider only what the parties want, but must also heed public perceptions of accuracy and fairness. As Schulhofer points out, justice and punishment are classic public goods. Allowing innocent defendants to plead guilty imposes "serious negative externalities" on society.⁸⁶ Society has a strong interest in ensuring not only that criminal convictions are in fact just, but also that they are perceived as just. Though one lawyer whom I interviewed cynically suggested that criminal justice is not and should not be about the truth, the public cares a great deal about truth.

Alschuler, however, turns this concern for public perceptions on its head. He praises Alford pleas as an "hones[t]" way to avoid "hypocrisy" instead of tempting innocent defendants to confess falsely.⁸⁷ But Alschuler's own evidence shows that many lawyers and judges are deeply uncomfortable with this prospect.⁸⁸ The public may be all the more uncomfortable, as Alschuler recognizes when he characterizes refusals to accept Alford pleas as a "public relations measure."⁸⁹ Alschuler even hints at a revolutionary goal of fomenting the overthrow of plea bargaining by exposing its internal contradictions.⁹⁰ The much more likely outcome, however, is maintenance of the status quo coupled with growing public cynicism about the entire system. Alford and nolo contendere pleas send mixed messages, which may breed public doubt, uncertainty, and lack of respect. Far from being honest, they let guilty defendants cloak their pleas in dishonesty. Recall that point from my interviews: most defendants who enter Alford and nolo pleas are guilty, so their protestations of innocence are false. Recall judges' and prosecutors' fear that defendants will enter Alford or nolo pleas and protest their innocence, undermining public trust in the justice system. Think of how the Sara Jane Olson case would have been perceived if Olson had pleaded without ever admitting guilt. In contrast, jury verdicts and unequivocal guilty pleas suppress residual doubts and promote public confidence.⁹¹

The justice system should forestall public cynicism by forbidding practices that openly promote injustice or even doubts about guilt. As the Supreme Court noted in Winship, the law goes to great lengths to minimize the risk of erroneous convictions. The perception of accuracy is needed "to command the respect and confidence of the community . . . . It is critical that the moral force of the criminal law not be diluted by a standard of proof that leaves people in doubt whether innocent men are being condemned."⁹² In other words, the system must not only be accurate to protect innocent defendants from injustice, but must also be seen as accurate and fair to secure the community's confidence.

Public confidence and faith in the justice system are essential to the law's democratic legitimacy, moral force, and popular obedience.⁹³ When citizens learn that defendants are pleading and being punished while refusing to admit guilt or even protesting their innocence, they may well suspect coercion and injustice.⁹⁴ These citizens may conclude that our system does not care enough about separating guilty from innocent defendants. Some may believe that the defendant is guilty but refuses to admit it, while others may doubt guilt and blame the system's callousness. A system less obsessed with efficiency would slow down and take a closer look at these cases, and the inefficient safeguards of trial might catch some of these injustices. But our obsession with efficiency and autonomy has led us to downplay the importance of justice and the public's perception of justice. This may partially explain why only one-third of the American public expresses confidence in the criminal justice system and why two-thirds think plea bargaining is a problem.⁹⁵

Because there are no published poll data on Alford and nolo contendere pleas, I conducted my own poll of law students. 62% of respondents doubted the fairness of hypothetical nolo contendere convictions, and 78% doubted the fairness of hypothetical Alford convictions. In addition, nolo contendere pleas led 27% of respondents to have less faith in the criminal justice system. Alford pleas led 51% to have less faith in the justice system. In short, a majority of respondents were troubled by these pleas.⁹⁶

Though this small poll is not definitive, it is highly suggestive. While proving a causal link between particular pleas and the justice system's reputation is usually impossible, prudence counsels erring on the side of caution. A serious concern for safeguarding innocent defendants, justice, and the popular perception of justice would support abolishing or at least severely restricting Alford and nolo pleas.

IV. VALUES OF THE SUBSTANTIVE CRIMINAL LAW

The preceding part opposed Alford and nolo contendere pleas based on their actual and perceived inaccuracy and unfairness. Though this conclusion is at odds with the scholarly literature supporting these pleas, this analysis rested on conventional procedural values. (It also added a moral dimension rarely seen in procedural literature, a clue that criminal procedure is detached from the moral moorings of the substantive criminal law). But what if we could ensure that defendants entered these pleas freely, without coercion or misunderstanding? What if we could make Alford and nolo contendere pleas perfectly accurate, say by requiring a stringent factual basis? And what if we publicized these factual bases enough to dispel popular concerns that innocent defendants are being convicted? Would we then want to let guilty defendants enter Alford and nolo contendere pleas, as an efficient, accurate, fair alternative to trial?

No. Even so, there would still be strong reasons to forbid these pleas. Even if these pleas were perfectly accurate and were seen as accurate and fair, they would undercut important values and norms of the substantive criminal law. Alford and nolo contendere pleas interfere with contrition, education, healing, and reform of defendants. And regardless of whether defendants learn these lessons, guilty-but-not-guilty pleas send muddied messages that obstruct catharsis and vindication of social norms and victims. These pleas undercut the norms of honesty and responsibility for one's actions by letting guilty defendants dishonestly dodge responsibility. Procedural efficiency does not justify undercutting these important substantive values, because substance is the very raison d'etre of procedure. Though substantive values need not trump procedural values, they should at least carry significant weight. Procedures that undercut these significant substantive values in effect cannibalize the law.⁹⁷

Part IV's overt examination of the moral messages sent by criminal procedures will make some readers uncomfortable. Other readers will disagree with my weighing of the moral pros and cons, if they put more value on defendants' privacy and less on reform, victim vindication, and expression of community condemnation. Criminal-justice discourse often avoids these objections by shying away from contentious moral disagreements. It seems easier to avoid morality and achieve consensus on seemingly apolitical issues of efficiency, accuracy, deterrence, and the like.⁹⁸ Nonetheless, moral questions lie at the root of criminal procedure. In practice, people's judgments about criminal justice hinge not on technical issues, but on the social meanings and moral judgments expressed by various practices.⁹⁹ These social, moral dimensions are central to evaluating criminal procedures; ignoring them will not make them go away.

Section A reviews the justifications for punishment that undergird the substantive criminal law. In addition to deterring and incapacitating, the criminal law also aspires to reform, educate, vindicate victims, produce catharsis, and express condemnation. Section A also reviews the basic moral norms embodied in the criminal law, such as honesty and responsibility for one's actions. Section B then explores the psychological denial mechanisms of offenders (especially sex offenders) who refuse to admit guilt. It also discusses the therapeutic value of inducing confessions by encouraging and confronting defendants who refuse to admit guilt. Section C contends that guilty pleas should be reserved for defendants who confess, because by confessing they take the first steps toward repentance and reform. It also explains why defendants who refuse to admit guilt should have to undergo trials. Trials may break through denial mechanisms and teach and heal offenders and victims. Even if defiant defendants refuse to repent, trials serve as morality plays that condemn crimes and vindicate victims and the community's norms. Conversely, trials vindicate innocent defendants, freeing them from the community's blame and censure.

A. Justifications for Punishment and the Law's Moral Norms

One influential strand of criminal law and scholarship takes a narrow, utilitarian view of the pains and pleasures associated with crimes. According to Jeremy Bentham, criminals commit crimes because crimes benefit them. To counteract these benefits, the criminal law incapacitates and deters by making the punishment attached to crimes sufficiently unpleasant and restrictive.¹⁰⁰

The criminal law does operate in part on this simple level of pleasure and pain. But it serves other, more morally laden functions as well. As Kant explained, punishment "ought to be done in order that every one may realize the desert of his deeds."¹⁰¹ The word realize has two relevant meanings, and Kant appears to mean both of them. First, Kant meant that offenders realize punishments the way that entrepreneurs realize profits: they reap what they have sown, the retribution that they have earned.¹⁰² Second, Kant also suggested that punishment makes us realize in our minds and hearts the wrongfulness of crimes. Punishment is a powerful "symbol" of moral blameworthiness that is "medicinal for the criminal and [sets] an example for others."¹⁰³ On this view, punishment reforms and deters in part by educating the offender and others.

The late Jean Hampton also espoused the theory of punishment as moral education. Punishment teaches the offender that the crime is forbidden because it is morally and legally wrong.¹⁰⁴ It is a strong tool for penetrating callous hearts. In C.S. Lewis's memorable phrase, it shatters our illusions and "plants the flag of truth within the fortress of a rebel soul."¹⁰⁵ It seeks to teach by triggering and developing the offender's moral sense of guilt. It tries to induce contrition and repentance, so that the offender will self-consciously repudiate his past wrongful act and feel guilty enough to avoid committing it again. ¹⁰⁶ As R.A. Duff puts it, punishment tries "to bring the criminal to understand the nature and implications of her crime; to repent that crime and thus, by willing her own punishment as a penance which can expiate her crime, to reconcile herself with the Right and with her community."¹⁰⁷ In a similar vein, Stephen Garvey sees punishment as a secular version of atonement, a way of reconciling offenders with victims and reintegrating them back into the community.¹⁰⁸ Before offenders can atone and be reconciled, they must first accept responsibility, learn their lessons, and resolve to mend their ways. ¹⁰⁹ True, some offenders will learn these lessons only in part, and some not at all. Nonetheless, the law respects their moral agency by at least trying to teach them the errors of their ways. Not every offender heeds this lesson, but the hope is that punishment increases the chances that offenders will repent and change their ways. This idea of repentance and atonement is all too often overlooked in discussions of the justifications of punishment.

Regardless of whether offenders learn their lessons and repent, punishment has moral value for others. Victims, for example, are demeaned by the crimes they suffer, which disregard and trample on their moral worth. Punishing offenders vindicates the worth of their victims, humbling wrongdoers by asserting that they are not entitled to abuse others as they like.¹¹⁰ It thus serves a cathartic function for victims and brings them closure. If wrongdoers confess-or, better yet, repent and apologize-victims can more readily forgive, surrender resentments, and find peace.¹¹¹ Some might understand the catharsis referred to in the text as simply satisfying the raw primal urge to punish, which arguably is an illegitimate goal of punishment. One can, however, instead understand it as in part a nobler satisfaction at seeing the law do justice and vindicate rules of right and wrong.

This symbolic moral significance of punishment extends beyond the victim to society at large. Punishment expresses condemnation, denouncing the wrong and reaffirming society's moral teaching of what is right.¹¹² As James Fitzjames Stephen put it, "the sentence of the law is to the moral sentiment of the public in relation to any offence what a seal is to hot wax. It converts into a permanent final judgment what might otherwise be a transient sentiment."¹¹³ In the process, our criminal procedures allow the community to vent its outrage, satisfying the public's sense of justice by bringing catharsis and closure.¹¹⁴

Though the literature often speaks of vindicating the community's norms, it rarely spells out what those moral norms are. Here is a sampling of the most prominent ones: The criminal law promotes honesty, good faith, and trustworthiness by stigmatizing perjury, fraud, and the like. It encourages self-discipline, peaceful conduct, and respect for persons and property by forbidding violent and property crimes. As a rule, it holds people responsible for their actions, treating them as moral agents. Yet it insists on moral culpability, in the form of mens rea and voluntariness, and excuses defendants who acted under duress, insanity, infancy, or other incapacity. Intent, blameworthiness, honesty, self-discipline, responsibility for one's actions, avoidance of violence, and protection of property-these are the most basic norms of the criminal law.

Of course, criminal law does not live up to its ideals. Our system does far less to educate, reform, reconcile, vindicate, and denounce than it could. There is an important gap between our ideals and reality, but the solution is not to abandon our ideals. Rather, where feasible, we must start to bring practice into harmony with those ideals. One has to start somewhere.

In short, criminal punishment does much more than deter and incapacitate as cheaply and swiftly as possible. It seeks also to educate the offender, induce repentance and reconciliation, vindicate the victim, achieve catharsis, and reinforce moral norms to society at large. Deterrence, education, repentance, and reconciliation are contingent on whether offenders learn their lessons and change their ways. In contrast, vindication of victims and moral norms are analytically tied to the very message punishment sends. Section B explains how confessions help to teach and change offenders in general and sex offenders in particular.

B. Reluctance to Confess and the Value of Confession

To achieve these goals, the criminal law seeks to a) lead offenders to repent by humbling them, b) exact moral sanctions, and then c) return them to the community as equals. The difficulty, as discussed later in this section, is that offenders cannot accept responsibility and repent until they own up to their actions. But admitting wrongdoing to oneself and to others is not easy. Many offenders are in denial about their wrongs. They may feel guilty inwardly about what they have done, and they may be ashamed to admit their wrongs to others.¹¹⁵ To avoid responsibility, they may publicly or privately deny their acts, deny their awareness, justify or excuse their conduct, or minimize its gravity or harm. Their inward and outward lies shield offenders from facing the painful truth.

These problems are most acute in the case of especially heinous or shameful crimes, such as sex offenses.¹¹⁶ It is no coincidence that sex offenders are among the most frequent users of Alford and nolo contendere pleas.¹¹⁷ Denial and minimization of culpability are the norm among sex offenders.¹¹⁸ Sex offenders deny the facts, their acts, their awareness, or their responsibility, or minimize the wrongfulness or impact of their behavior.¹¹⁹ They deny to their families, friends, employers, and society at large to avoid shame and embarrassment.¹²⁰ They are also "afraid to admit the truth, even to themselves. The thought of being a sexual deviate can be so frightening or repugnant to them that they hide from themselves for years."¹²¹ In short, sex offenders lie to avoid or minimize shame and guilt. (In contrast, white-collar defendants are more likely to lie to avoid shame and protect their reputations.¹²²)

In many cases, these lies and explanations are not simply excuses for public consumption. They reveal underlying attitudes and cognitive distortions that may lead to more sexual offenses in the future. ¹²³ Offenders who lie to others come to lie to themselves and distort their memories and interpretations of events.¹²⁴ For example, a molester might say "It wasn't sexual and anyway there's nothing wrong with having sex with a child," or a rapist might say and believe that the victim asked for it.¹²⁵

These cognitive distortions and denials impede treatment. Admitting one's wrongdoing is the first step toward moving beyond it.¹²⁶ In twelve-step programs such as Alcoholics Anonymous, an essential step is to admit that one has a problem.¹²⁷ Confessed offenders can no longer rest complacent in the illusions that they are good people, but must admit that they have fallen down as a prelude to standing up again. In addition, denials prevent offenders from revealing details of their offenses, which are essential to framing a therapeutic response.¹²⁸ They prevent therapists from examining cognitive distortions, detecting warning signs, and nurturing empathy for victims.¹²⁹ Thus, most treatment programs refuse to admit sex offenders who deny any sexual conduct.¹³⁰ According to my interviews, the sex-offender treatment program in Missouri state prison requires admission of guilt as a condition of therapy. Thus, Missouri judges will not allow sex offenders to enter Alford pleas. Denial, in short, obstructs treatment, which in turn greatly increases the risk of recidivism.¹³¹

When wrongdoers are reluctant to admit responsibility, many therapists find it helpful to confront them with the facts, to force them to come to terms with their behavior.¹³² Firmly challenging these denials and distortions is a "very effective" way of overcoming them. ¹³³ Therapists must actively confront and challenge sex offenders; supportive, passive therapy usually fails.¹³⁴ Challenges can take various forms, including group confrontation, questioning of inconsistencies, asking for explanations and details, indirect challenges, challenges by loved ones, challenges coupled with empathy and support, and challenging offenders to challenge themselves.¹³⁵ By challenging offenders' excuses and rationalizations, therapists can trigger feelings of guilt and harness this guilt to induce change.¹³⁶ Even external pressures, such as the threat of imprisonment, can be effective inducements to overcome denial.¹³⁷ These external challenges and pressures help to shatter the offender's illusions and begin the journey.

Confessions and denials within the legal system may have effects similar to confessions and denials within therapy. Confessions in open court, even if induced by some external pressure, can begin to breach the dam of denial. If offenders who confess later try to recant during treatment, therapists can confront them with the details of their initial confessions.¹³⁸ In contrast, repeated unchallenged denials in the legal system only exacerbate offenders' denial reflex, making subsequent treatment even harder.¹³⁹ The result is that offenders who enter Alford or nolo pleas remain in denial and resist successful treatment, making them much more likely to reoffend.¹⁴⁰ One small Minnesota study found that seven out of eight sex offenders who had entered Alford pleas reoffended within five years of release. ¹⁴¹ This is two to five times the recidivism rate of sex offenders in general.¹⁴²

Two anecdotal interviews with judges indicate that what offenders say in court matters to their and others' perceptions of their guilt. One long-time judge reported to me that he used to allow defendants to plead nolo contendere. What he found was that a defendant would say nothing in court. Then, upon reporting to a probation officer for a pre-sentence interview, he would deny guilt. He would also tell his family that he had not done the crime but his lawyer had made him plead guilty. Family members would write angry letters to the judge, complaining about what a travesty of justice it was to convict an innocent man. They would say, for example, that the rape victim was a tramp who consented to sex. At sentencing, the defendant and his family would continue to deny guilt and, at least implicitly, blame the victim. Victims would be visibly frustrated when making statements at sentencing, feeling that they had to justify themselves instead of being able to heal. After sentencing, these convicted defendants would continue to deny guilt, impeding therapy or treatment. Then, this judge stopped permitting most nolo contendere pleas. Defense lawyers had to confront clients and bring them to admit guilt; almost none insisted on going to trial. They did not renege on these admissions after leaving court. Defendants evidently stopped denying guilt to themselves and their families, as the complaining letters from families stopped. Defendants and families no longer denied guilt at sentencing; they seemed less defiant, more contrite, and less openly hostile and angry. Victims took satisfaction in feeling vindicated and were able to express healthy outrage at sentencing instead of frustration. To this judge, the victims appeared to be healing better. And the judge, having heard a detailed plea colloquy, was much better able to confront defendants with the details and wrongfulness of their acts.¹⁴³ Another judge confirmed the outlines of this story. He noted that some defendants are agitated and balk at admitting guilt, but plead guilty when told that the alternative is going to trial. These defendants seem calmer and more accepting of responsibility after their guilty-plea allocutions and less likely to protest innocence and injustice later on. In short, the legal system's labels and unequivocal ascriptions of blame influence offenders' thought processes and their treatment prospects.

An analogous dynamic may be at work in insanity cases. A few case studies show that offenders who are found not guilty by reason of insanity resist discussing their thoughts, feelings, and actions. Instead, they externalize their feelings and ascriptions of blame. They may show no remorse, saying (as two patients did) "The judge said I was not guilty" or "I have not committed a crime."¹⁴⁴ In contrast, persons with mental illness who are convicted of crimes may react more positively. Society's pronouncement of guilt may spur and reinforce the offender's introspection, acceptance of responsibility, and treatment prospects.¹⁴⁵ In short, "judicial expression[s] of blameworthiness" promote "accept[ance of] responsibility for actions committed during periods of gross mental disorder," which may aid treatment and reform. ¹⁴⁶ My point is not that the insanity defense is good or bad-there are many other pros and cons to consider. My point is simply that the law's labels and ascriptions of blame influence rehabilitation and reform.

This evidence comes primarily from the psychological literature on sex offenders, substance abusers, and mentally ill offenders. I have found only one study that discusses the psychology of guilt and confessions in criminals generally. That study confirms that guilt, psychological blocks, and confessions play similar roles in treating other kinds of offenders.¹⁴⁷ Perhaps it is dangerous to generalize, but one might extrapolate based on my interviews that other offenders who enter Alford and nolo contendere pleas are doing so in part because they face similar psychological blocks. Alschuler's anecdotal evidence certainly suggests that.¹⁴⁸ The idea is intuitively plausible. Offenders who are not reluctant to confess enter straight guilty pleas. Offenders whose psychological barriers impede confession (to others or even to themselves) are the prime customers for Alford and nolo pleas, as my interviews confirmed.¹⁴⁹ Those guilty defendants who use these pleas most are presumably those in the deepest denial who most need to come clean.

Some defendants are willing to confess and plead guilty. As Alschuler rightly notes, most guilty pleas are not the fruit of genuine repentance. Instead, defendants feign repentance to earn sentence reductions.¹⁵⁰ But even feigned or induced repentance may teach lessons to some offenders. The very act of confessing and pleading guilty in open court drives home the defendant's awareness of the victim's injury, the norm violated, and the community's condemnation.¹⁵¹ Indeed, the ordeal of feigning repentance, even if done initially for the wrong reasons, can sometimes lead to genuine repentance.¹⁵² Cognitive psychology teaches the same point. According to cognitive dissonance theory, persons who publicly take positions that they do not believe are likely to change their attitudes to bring them into line with their public statements. Kenneth S. Bordens & Irwin A. Horowitz, Social Psychology 221 (2002). Thus, offenders who publicly accept responsibility for their crimes, even if they do so insincerely, are more likely to internalize that responsibility than those who persist in denying guilt. Confessing is for many a difficult act. It requires owning up to shameful deeds, putting aside excuses and evasions, and taking responsibility for one's past. As my interviews indicated, defense lawyers often have to work with defendants before they can bring themselves to admit guilt. The hard work of admitting guilt and repenting may impress upon the defendant the wrongfulness and gravity of the crime. By admitting guilt, however insincerely, defendants let down their denial mechanisms, open the path to reform, and bring closure to the community.¹⁵³ The value of confronting guilt and apologizing is a central insight of the restorative-justice movement. Offenders confront their guilt and wrongdoing by meeting face-to-face with victims, learning about their sufferings, and perhaps apologizing and making amends. See generally John Braithwaite, Restorative Justice & Responsive Regulation (2002); Gerry Johnstone, Restorative Justice: Ideas, Values, Debates (2002); Garvey, supra note , at 1840-44. Doubtless many offenders start out making grudging or insincere admissions and apologies. Nonetheless, the hope is that victim-offender mediation, sentencing circles, family-group conferences and the like will teach offenders moral lessons by making them see the suffering they have caused.

Perhaps many defendants plead guilty cavalierly, confessing the words without confronting their significance. But this description is least true of those defendants who balk most at pleading guilty, namely those who want Alford and nolo pleas. These defendants are in the deepest denial and would have to struggle the most to admit guilt. This struggle to admit shameful acts is more likely to leave an impression on them, because it involves breaking down walls of denial. The more of a struggle it is to confess, the bigger the breakthrough when the defendant finally confesses. Indeed, it is a catharsis, literally a cleansing, which is why we often speak of confession as coming clean.

In short, confessions help to educate and reform some defendants. And by accepting responsibility, guilty defendants who confess bear witness to the truth and bring closure to victims and the community.

C. The Substantive Value of Trials

Whatever their other flaws, plea bargains induce guilty defendants to confess and start down the road to repentance. Some defendants cannot or will not admit guilt. For these guilty defendants, as well as for innocent defendants, the law has jury trials.¹⁵⁴ We usually think of jury trials as simply procedural safeguards designed to ensure accuracy and fairness. To borrow Herbert Packer's terminology, there are two dominant approaches to criminal procedure. Crime-control advocates stress speedy and efficient pursuit of the truth; due-process advocates emphasize procedural fairness and perceived fairness.¹⁵⁵

Both views of criminal procedure, however, are incomplete. Trials not only seek fairness, efficiency, and accuracy; they also further the criminal law's substantive moral aims and norms. As Thurman Arnold put it: "Trials are like the miracle or morality plays of ancient times. They dramatically present the conflicting moral values of a community in a way that could not be done by logical formalization."¹⁵⁶ The jury serves as the chorus of a Greek tragedy, "the conscience of the community."¹⁵⁷ It applies the community's moral code, pronounces judgment, and brands or exonerates the defendant.¹⁵⁸ The parade of live witnesses and the solemn pronouncement of guilt confront the offender at length with his wrongful deeds. This litany of accusation, evidence, and condemnation may perhaps break through the defendant's denial mechanisms, driving home in undeniable detail the wrongfulness of the crime. (Alternatively, the preparations for and prospect of facing trials may force offenders and counsel to confront guilt, eventually leading to straight guilty pleas.) These morality plays hold out hope of reforming guilty defendants and healing society. Colonial Americans prized the trial as "an occasion for repentance and reintegration: a ritual for reclaiming lost sheep and restoring them to the flock."¹⁵⁹

For those offenders who refuse to confess or repent, trials still bring catharsis and closure to victims and the community. As one court put it, "[j]ury trials have historically served to vent community pressures and passions. As the lid of a tea kettle releases steam, jury trials in criminal cases allow peaceful expression of community outrage at arbitrary government or vicious criminal acts."¹⁶⁰ Or, as the Supreme Court put it, "public trials ha[ve] significant community therapeutic value" and bring "community catharsis."¹⁶¹ Trials express respect for the law, communicate values, and justify punishment. They engage offenders in critically examining their acts, and they express and justify the community's norms and response.¹⁶² On Durkheim's view, convictions at trial vindicate victims and the community, denouncing offenders and reaffirming moral norms in the face of their transgression.¹⁶³ This is true regardless of how offenders respond.

Conversely, acquittals at trial vindicate innocent defendants and the moral norms on which they acted.¹⁶⁴ Think, for example, of the prosecution of John Peter Zenger for seditious libel in colonial America. The jury's celebrated acquittal proclaimed to all eternity Zenger's right to criticize the government.

Trial verdicts, whether convictions or acquittals, send messages about the substantive moral norms underlying the criminal law. The message of one case is that a bank clerk, no matter how tempted, must restrain herself from helping herself to other people's money. Another case teaches that hate crimes and bigotry against minorities are wrong. The message of a third case condemns the impulsive killer's failure to restrain himself in the face of a mere insult. Each unambiguous verdict is a lesson vindicating a moral norm of the community. Some trials are widely publicized, while others speak only to the victims, witnesses, jurors, and others directly involved. Though the scope of publicity varies, each trial sends a message to the universe of those who hear about it.

If trials serve these purposes, then Alford and nolo contendere pleas subvert these purposes in the name of efficiency and autonomy. Convictions at trial and ordinary guilty pleas communicate unambiguous verdicts of guilt that bring closure.¹⁶⁵ This clarity is essential. Alford and to a lesser extent nolo contendere pleas are ambiguous on their faces. By pleading guilty-but-not-guilty, defendants muddy the moral message. These pleas communicate that the law does not care enough about crime to insist upon clear, honest resolutions and vindications. Proclaiming the truth, justice, respect for others, and self-restraint take a back seat to procedural efficiency and freedom of choice. Plea bargains may do that covertly, but Alford and nolo contendere pleas do it shamelessly.¹⁶⁶ By failing to challenge offenders who falsely deny guilt, criminal procedure undermines the criminal law's basic norms of honesty and responsibility for one's actions.

Some authors would rather shelter guilty defendants from the norms of justice. As Kaden recognizes, Alford and nolo contendere pleas allow defendants to preserve their autonomy, privacy, and dignity by refusing to admit guilt.¹⁶⁷ Unfortunately, this autonomy, privacy, and dignity come at the expense of education, repentance, reconciliation, and vindication. Offenders have abused their autonomy and privacy, forfeiting their right to live autonomously and privately outside prison walls. They have exalted their freedom at the expense of victims and need to humbly accept responsibility for their wrongdoing.¹⁶⁸ The Psalmist ties this humbling to the law's function of educating wrongdoers: "It is good for me that Thou hast humbled me, that I might learn Thy statutes." The Psalter According to the Seventy Psalm 118:71 (Holy Transfiguration Monastery trans. 1997) (compare Psalm 119:71 (King James Version): "[It is] good for me that I have been afflicted, that I might learn thy statutes."). Herbert Morris notes that offenders have aggrandized themselves at the expense of victims by renouncing compliance with the law.¹⁶⁹ Punishment seeks to "humbl[e] the [defendant's] will," "to bring him low" after he has aggrandized himself.¹⁷⁰ A guilty plea allocution or a full-fledged jury trial teaches this lesson to the offender, the victim, and the community. A quick Alford or nolo contendere plea short-circuits the process, leaving offenders and their families in denial.¹⁷¹ Offenders avoid the shame and guilt of owning up to their deeds, though they do not deserve to avoid the shame and guilt that they have earned. Their families continue to protest innocence, instead of reinforcing the message that it is time to change. These pleas are most attractive to those who deny guilt to others and perhaps themselves, but these are precisely the offenders who need to have their illusions, excuses, and pride shattered. If offenders do not admit to having done anything wrong, they are unlikely to learn the lessons of their crimes and move beyond them. As Santayana warned, "those who cannot remember the past are doomed to repeat it."¹⁷² These are the offenders who most need to learn their lessons at jury trials.

Another justification offered for these pleas is that they ease the strain on the relationship between defendants and defense counsel. As Alschuler and Wallburn suggest, defendants often balk at admitting guilt, even to their lawyers. Defense lawyers can avoid friction by allowing clients to enter Alford pleas instead of pressing clients to admit guilt.¹⁷³ So, their reasoning goes, these pleas avoid straining the attorney-client relationship. This view fits with the dominant view of lawyering as gamesmanship. On this view, the job of a defense lawyer is to avoid conviction, minimize punishment, and further whatever interests the client wishes to pursue. There is a spectrum of ways to approach serving the client's interest. At one end of the spectrum, some lawyers actively confront and challenge their clients when preparing for trial or seeking leniency at sentencing. Toward the other end of the spectrum, some lawyers are more hands-off about challenging clients' assertions of fact or expressions of preference.

Allowing lawyers to take this path of least resistance is a vice, however, not a virtue. This is not the place to consider the merits of gamesmanship more generally. But at least in the context of defendants who are in denial, the gamesmanship model is misguided. It ignores the constructive role that defense lawyers can play in educating and transforming clients' misperceptions and short-term desires. Instead, it takes short-term desires as a given, even when these clients are suffering from psychological blocks that obstruct their long-term interests and values.¹⁷⁴ (In contrast, guardians are often authorized to put their wards' long-term interests above their short-term desires.)¹⁷⁵ True, the Model Rules of Professional Conduct distinguish between the roles of guardians and lawyers. They provide that, to the extent possible, lawyers should treat disabled clients just like any other clients. Model Rules of Professional Conduct Rule 1.14. But they also recognize that in some situations, wards need guardians to make decisions for them or at least guide them in making these decisions. The proposal in the text would still allow clients to make ultimate decisions, but would constrict the range of legally permissible decisions in order to foster more productive confrontation and dialogue. Lawyers can recognize that substance abuse, mental illness, psychological blind spots and denial, or simple short-sightedness impedes their clients' rationality.¹⁷⁶ They can persuade clients to face up to patterns of behavior that, if left unattended, will produce more crimes and punishment in the long run.¹⁷⁷ As suggested by the psychological literature cited earlier, lawyers can confront their clients with the overwhelming evidence of guilt and break down their denials.¹⁷⁸ And lawyers can provide moral as well as legal counsel, advising clients that it is right to come clean, apologize to victims, and move forward.¹⁷⁹ By penetrating clients' denials to others and themselves, defense counsel can begin the process of honesty, education, and reform.

Some defense attorneys do challenge their clients, but not all do. Defense lawyers can instead say, "Yes, you are innocent, but a jury would probably convict you at trial, so enter an Alford or nolo contendere plea." As one psychologist notes, defense lawyers exacerbate the problem by failing to challenge their clients' denials enough.¹⁸⁰ Far from challenging denial and encouraging confession, the dominant client-centered approach to counseling discourages painful confession.¹⁸¹ Indeed, some defense lawyers purposely avoid learning all the facts about guilt, so that they remain free to make arguments that run counter to the undiscovered facts.¹⁸² This see-no-evil approach not only leaves offenders' illusions and denials in place but compounds them. Clients interpret this failure to challenge as confirmation and become even more resistant to the challenges required during therapy and rehabilitation.¹⁸³ Instead of repenting and moving forward, offenders may continue to feel sorry for themselves, focusing their energies on collateral attacks instead of making amends. The criminal process is not simply about getting to prison as smoothly as possible, but also about teaching and opening the offender's eyes along the way. (If on the other hand the client is innocent, this approach is a cop-out that avoids vigorously vindicating innocence.) In short, letting defense counsel take the easy way out actually disserves the client's long-run interests, as it gets in the way of facing the facts. The same is true of judges and prosecutors who acquiesce in Alford or nolo contendere pleas. They may be serving their own selfish or short-sighted interests in clearing their dockets instead of seeking to do more to heal, teach, and vindicate.

Regardless of whether Alford and nolo contendere pleas impair the offender's own education and reform, they hurt others. They undercut not only the lessons taught by the criminal justice system, but also its deterrent effect on this and future offenders. Offenders dislike having to admit their guilt, and this dislike makes committing a crime marginally more unpleasant. Offenders also dislike being collaterally estopped in related civil lawsuits, which again should have an additional deterrent effect.¹⁸⁴ Alford pleas allow offenders to avoid the former consequence. Nolo contendere pleas let them avoid both, which serves the self-interests of offenders and prosecutors at the expense of victims and the civil courts. Victims already find it hard enough to collect compensation and restitution even with the benefit of estoppel. Nolo contendere pleas compound the problem by not estopping offenders, instead letting them relitigate their convictions. Victims and civil courts must spend more time and money to collect compensation. Yet there is no good reason to allow relitigation: if there is proof beyond a reasonable doubt of a crime, then a fortiori there is proof by a preponderance of the evidence, and estoppel should attach.¹⁸⁵ It is hard to know exactly how significant estoppel is in practice or how often it becomes an issue in subsequent civil litigation. My interviews revealed that it was a frequently mentioned justification for nolo contendere pleas, which implies that the prospect of civil litigation comes up with some regularity. This is particularly true in automobile-accident cases (because insurance companies with deep pockets are involved). See supra Part II.C. The failure to estop defendants and drive home guilt is yet another way that these pleas equivocate and undermine the criminal law's message. By alleviating the unpleasantness of convictions, these pleas weaken general and specific deterrence.

Alford and nolo contendere pleas also hurt victims and the community. Victims go without vindication. They lose their day in court, their chance to vent their sorrows and ask for justice, without even an admission of wrongdoing let alone an apology.¹⁸⁶ (This explains why, when victims find out about Alford pleas ahead of time, they sometimes try to dissuade prosecutors from accepting them.¹⁸⁷) Molestation victims, for example, can suffer more harm when courts appear to accept the denials of molesters; this judicial acceptance seems to suggest that the victims are liars.¹⁸⁸ And the lack of estoppel makes it harder for victims to collect from defendants who pleaded nolo contendere. Traumatized victims who want closure may be more reluctant to pursue these claims at all. Society loses the authoritative vindication of its norms and the repudiation of the wrong, from either the jury or the offender. Instead of communicating that punishment is moral denunciation based on true desert, society treats it as a marketable good, undermining its moral authority.¹⁸⁹ Of course, many pleas are not well publicized, but to the extent that victims and others do learn of them, they take away the wrong message. In short, the efficiency of these pleas comes at the price of undermining catharsis, expressive condemnation, and vindication of the community's norms.

Alford and nolo contendere pleas may well be constitutional, in light of the latter's historical pedigree.¹⁹⁰ But constitutionality is not wisdom. Though our Constitution may tolerate these pleas, they are nonetheless unwise and should be abolished. Indeed, Alford left this door open, recognizing that while these pleas are permissible they are not required.¹⁹¹ Legislatures are perfectly free to ban these pleas. Until they do so, judges are free to exercise their discretion to reject these pleas. Prosecutors may vigorously oppose them, and defense counsel may advise their clients to confront guilt instead of taking the easy way out. (Perhaps no-contest pleas are necessary in isolated cases, such as where a defendant was truly too intoxicated to remember anything and a trial would traumatize a vulnerable witness. This kind of justification, however, should be exceptional. They might also be needed in traffic cases, where a tort suit will resolve the fault issue and the minor traffic violation carries less moral significance. It is hard to imagine any case in which an Alford plea would be needed, particularly because Alford pleas are far more ambiguous and equivocal.) Unequivocal pleas or trials may be less efficient and require more work, but in the long run they better serve the moral values of the criminal law.

CONCLUSION

Our guilty plea process has become streamlined, to allow the maximum entry of guilty pleas with a minimum amount of work. It is so obsessed with efficiency that even a defendant's refusal to admit guilt or protestations of innocence do not give us pause. This assembly-line approach to criminal procedure downplays the importance of innocence and fairness, but it also ignores the important substantive values of the criminal law. These values are not even on the radar screens of proceduralists, when in fact they should take precedence. Procedure is supposed to further substance instead of cannibalizing it. Scholars should closely scrutinize procedures that may undercut substantive values, but they do not.

This problem is not an isolated one. It reflects a broader gulf between procedure and substance. Scholars often reify the artificial curricular distinctions between subjects. This can obscure how one subject affects another, how procedure serves or hinders substance and vice versa. We often ask whether a given procedure is efficient, or accurate, or constitutional. Efficiency, in particular, has driven us to our wholesale embrace of guilty pleas, lest we be burdened with expensive and time-consuming trials. But we rarely ask if a procedure deters, educates, inflicts retribution, or produces catharsis. Nor do we ask if it undercuts the criminal law's moral norms, such as honesty, trustworthiness, self-discipline, non-violence, and respect for others.

If we were to look at procedures through this substantive-values lens, the disappearance of jury trials would be cause for more concern. Jury trials are far from efficient, and they are not always accurate either. But they serve as morality plays, expressing the community's conscience, allowing victims to voice grievances, and teaching offenders and others by vindicating society's norms. Jury trials have all but disappeared, accounting for fewer than 4% of all cases disposed of by trial or plea¹⁹², and wishing will not make them come back. But in the real world of guilty pleas, we can at least try to incorporate these values into our plea and sentencing procedures.

I suggested in Part IV that this substantive-values approach might require rethinking the role of defense counsel. Instead of taking their clients' desires as a given, defense counsel might try to educate and serve their long-term interests. (For example, defense counsel might do their clients a great service by pressing them to seek drug treatment.) Scholars should likewise rethink the roles of judges and prosecutors. The issue is not simply one of favoring defendants versus favoring prosecutors-that view buys into the stale dichotomy between crime-control and due-process approaches. It is time to transcend this zero-sum way of looking at criminal procedure, as efficiency versus fairness, and instead think about what other values prosecutors, judges, and defense lawyers should serve. The win-at-all-costs mentality that sometimes prevails at the bar might give way to a broader approach. Perhaps the right incentive structures could encourage prosecutors and judges to care less about maximizing convictions and case dispositions. Perhaps they could see the law and their roles more nobly: in terms of public morality, as they seek to teach, vindicate, heal, and reconcile.

For example, at guilty plea allocutions, defendants can choose to make full, honest, and open admissions of guilt and expression of remorse. But defendants often do not go nearly so far. Sometimes, they combine bare-bones or grudging admissions to crimes with excuses, evasions, or deflecting blame onto others. Currently, the federal Sentencing Guidelines are vague as to when guilty pleas will earn sentencing reductions for acceptance of responsibility.¹⁹³ In practice, however, many federal judges automatically award full acceptance-of-responsibility reductions for any guilty plea.¹⁹⁴ Many do so no matter how grudging the defendant's admissions were and no matter how little remorse he displayed.¹⁹⁵ This approach would make sense if our only concern was reaping the efficiency benefits of avoiding trials. But if moral education and reform are also important goals of the process, then judges should calibrate the rewards to the defendant's contrition and candor.

If defense lawyers knew that full acceptance of responsibility mattered, they would press their clients harder to come to terms with their misdeeds. Some defendants would still be grudging or insincere in admitting guilt, and these may be the ones who most need punishment to break through their half-denials. Even insincere apologies and promises of reform have value. For example, compliance theory suggests that defendants who publicly commit to reform in front of a judge and loved ones are much more likely to reform.¹⁹⁶ Judges should consider defendants' demeanors at pleas when imposing sentence, in assessing what lessons they need to learn. Thus, the judge who will impose sentence should hear the guilty plea whenever possible, instead of passing the plea off to a magistrate (as federal judges often do).¹⁹⁷

There are other ways to inject substantive values into plea and sentencing procedures. For example, judges at plea hearings could insist on more detailed allocutions and use more overtly moralistic language, driving home the wrongfulness of the crime. Victims could confront offenders at plea or sentencing, giving voice to the grief they have suffered. This may be the best argument for victims'-rights bills: They serve overlooked substantive values such as retribution, catharsis, and closure, and counteract prosecutors' temptations to sacrifice these values for efficiency.¹⁹⁸ And the rhetoric of sentencing could become less a desiccated recitation of Sentencing Guidelines mathematics and more an expressly moral judgment on wrongdoing. These and similar procedures would help to educate, express communal outrage, vindicate norms, and reform as jury trials once did.

There is a range of ways to apply this article's substantive-values approach. A modest view would simply add substantive values to the balance or mix of values that proceduralists currently consider. A strong view would say that these substantive values should become the dominant concern of criminal procedure. Depending on how far one takes it, this approach runs up against some entrenched features of our system. For example, an important strain of my argument has favored a more moral, didactic emphasis in proceduralist reasoning. One might use this approach to scrutinize the messages sent by many established procedures: exclusionary rules, plea bargaining, the adversary system, and even our rights-based approach. My point is not to resolve the tradeoff between substantive and procedural values, nor to discuss how much scholars must accommodate established practice. Procedural values, stare decisis concerns, and practical problems will all limit the implementation of substantive values. My point is simply to show the range of procedures that one could examine using this new lens.

Implementing these substantive values will not be simple, as they may be less convenient and might require more time and effort per case. But the cheapest and fastest methods are not always the best ones, and some things are worth the cost. The criminal law is not simply about locking up offenders cheaply.¹⁹⁹ If the criminal law is to educate the public and teach offenders lessons, we must be willing to spend money, just as we must spend for good educations in schools. Procedure must reinforce substance, instead of sacrificing it on the altar of short-sighted efficiency.

This bridging of the procedure/substance divide is part of a broader project to refocus criminal procedure. Criminal procedure scholarship has been so preoccupied with the constitutionality of rules that it has neglected their wisdom. And scholars focus so much on what the Supreme Court of the United States does that we ignore the rest of the actors in the process. The Constitution, however, sets only a floor, not a ceiling. The Supreme Court permits Alford pleas, but that does not mean that legislatures, prosecutors, defense lawyers, and trial judges should use them. It is time to move beyond our ivory-tower focus on the Supreme Court and constitutional law. We must also scrutinize how procedures do and should interact with substantive values in the real world.

*Associate Professor, University of Iowa College of Law; former Assistant U.S. Attorney, Criminal Division, U.S. Attorney's Office for the Southern District of New York (bibas@philo.org). B.A. Columbia; B.A., M.A. Oxford; J.D., Yale. Thanks to Al Alschuler, Dave Baldus, Anita Bernstein, Randy Bezanson, Peter Blanck, Steve Garvey, Gary Goodpaster, Herb Hovenkamp, Ken Kress, Margaret Raymond, Chris Slobogin, David Wexler, Bruce Winick, and Tung Yin and participants in faculty workshops at Northwestern University, the University of Iowa, University of Kansas, and University of St. Thomas Schools of Law for their advice and comments on earlier drafts and to Bryan Bennett, John Harbin, Keith Kasten, Ted Moore, and Kerstin Rademacher for able research assistance. © 2003 by Stephanos Bibas; all rights reserved.

FOOTNOTES

t1 See William J. Stuntz, The Uneasy Relationship Between Criminal Procedure and Criminal Justice, 107 Yale L.J. 1, 6 (1997) (noting that "[c]riminal procedure is, basically, a subset of constitutional law" and decrying the artificial separation between criminal procedure and substance to which this leads).

t2 Infra text accompanying notes -.

t3 Infra text accompanying notes -.

1 For the sake of brevity, I use "pleas" as a shorthand for pleas of nolo contendere and guilty (including Alford pleas).

2 Infra Part II.A.

3 Infra Part II.A.

4 Infra Part II.B.

5 Infra notes - and accompanying text.

6 Albert W. Alschuler, The Defense Attorney's Role in Plea Bargaining, 84 Yale L.J. 1179, 1294 (1975) [hereinafter Alschuler, The Defense Attorney's Role].

7 I am not arguing that substantive values should always trump procedural values, nor do I have a simplistic metric for prioritizing or balancing the two. My point is more modest: substantive values ought at least to factor into our vocabulary, our discussion, our consideration of procedures and procedural rules.

8 I recognize that many guilty-plea confessions are insincere or induced by extrinsic inducements or pressures, such as plea bargains. As I contend later in the article, however, even a true but insincere confession is better than no confession at all. It may help to break down the guilty defendant's denial mechanisms as the first step on the road to reform. Even if it does not, it teaches, heals, and vindicates the victim and society's moral norms. See infra Part IV.B-C.

9 Telephone interview with Shawn Chapman, Partner, Vorzimer, Masserman & Chapman and defense counsel to Sara Jane Olson (May 28, 2002); Telephone interview with Eleanor Hunter, Assistant District Attorney, County of Los Angeles and prosecutor of Sara Jane Olson (May 28, 2002); see also Declaration of Shawn Snider Chapman ии 34, 38, 39, 41, 42 (Nov. 5, 2001) (available on-line at http://www.saraolsondefense.com/Trial/Motions/declaration111001.html) (declaring that a prosecutor rebuffed the idea of a no-contest plea; also declaring that Olson initially refused to admit guilt and the prosecutor suggested that Olson would not have to do so, but later that day the District Attorney insisted that Olson would have to admit guilt as part of a plea agreement).

10 James Sterngold, Comments by 70's Radical Cast Doubt on Plea Deal, N.Y. Times, Nov. 2, 2001, at A12; James Sterngold, 70's Radical Pleads Guilty in Bomb Plot, N.Y. Times, Nov. 1, 2001, at A18.

11 James Sterngold, 70's Radical Reaffirms Guilty Plea, N.Y. Times, Nov. 7, 2001, at A16.

12 Id.; James Sterngold, Waffling Again, 70's Radical Asks to Change Guilty Plea, N.Y. Times, Nov. 15, 2001, at A20.

13 Change of Plea in Bomb Case Is Challenged, N.Y. Times, Nov. 30, 2001, at A20; James Sterngold, Waffling Again, 70's Radical Asks to Change Guilty Plea, N.Y. Times, Nov. 15, 2001, at A20.

14 James Sterngold, Judge Refuses to Permit Ex-Radical to Reverse Plea, N.Y. Times, Dec. 4, 2001, at A12.

15 James Sterngold, 70's Radical Is Sentenced, Then Arraigned in New Case, N.Y. Times, Jan. 19, 2002, at A10; James Sterngold, A Radical's Tale: Compassion Then Led to Prison Now, N.Y. Times, Dec. 14, 2001, at A24; see also John M. Broder, In a Quiet End to a Case, 4 Ex-Symbionese Liberation Army Members Plead Guilty to Murder, N.Y. Times, Nov. 8, 2002, at A14 (noting that Sara Jane Olson later pleaded guilty to a related murder, saying "I am truly sorry, and I will be sorry until the day I die").

16 Santobello v. New York, 404 U.S. 257, 260-61 (1971) (stating that plea bargaining "is not only an essential part of the process but a highly desirable part" because pleas save resources, are "prompt and final" dispositions, and start the correctional processes promptly); Brady v. United States, 397 U.S. 742, 752 (1970) (stressing that guilty pleas benefit both sides by avoiding the burdens and expenses of trial, reducing defendants' maximum punishment, and speeding up final dispositions and punishment); see also Warren Burger, The State of the Judiciary-1970, 56 A.B.A. J. 929, 931 (1970).

17 Santobello, 404 U.S. at 261-62.

18 Milton Heumann, Plea Bargaining 24-33, 156-57 (1978) (contending that the bench and bar like plea bargaining because it saves time and money, caps defendants' sentences, and expedites convictions that are almost inevitable, but disputing the hypothesis that caseload pressures explain plea bargaining); George Fisher, Plea Bargaining's Triumph, 109 Yale L.J. 857, 865-68, 893-903, 988-1001, 1074-75 (2000) (defending caseload pressure as an explanation for the growth of plea bargaining in Massachusetts).

19 Frank Easterbrook, Criminal Procedure as a Market System, 12 J. Legal Stud. 289, 308-09 (1983) (arguing that plea bargaining is a desirable mechanism for setting the price of crime); see also Thomas W. Church, Jr., In Defense of "Bargain Justice," 13 Law & Soc'y Rev. 509, 513-15 (1979) (arguing that defendants make rational decisions to plead or go to trial based on the expected sentence after trial discounted by the likelihood of acquittal).

20 Easterbrook, Criminal Procedure as a Market System, supra note , at 317 ("[T]he autonomy value-the right to waive one's rights as one method of exercising them- .. . underlies plea bargaining in this country."); Frank Easterbrook, Plea Bargaining as Compromise, 101 Yale L.J. 1969, 1975 (1992); Robert Scott & William Stuntz, Plea Bargaining as Contract, 101 Yale L.J. 1909, 1911-17, 1930-40 (1992) (treating expanded choice as a norm justifying a "presumption of enforceability"; also noting that parties enter plea bargains to exchange risks).

21 Scott & Stuntz, supra note , at 1918-35 (also noting dangers of substantively unconscionable or unequal outcomes, but arguing that abolishing plea bargains would be no better for defendants).

22 Albert W. Alschuler, Implementing the Criminal Defendant's Right to Trial: Alternatives to the Plea Bargaining System, 50 U. Chi. L. Rev. 931, 932-33 (1983) [hereinafter Alschuler, Implementing the Criminal Defendant's Right to Trial]; Albert W. Alschuler, The Changing Plea Bargaining Debate, 69 Calif. L. Rev. 652, 703-07 (1981); Albert W. Alschuler, The Trial Judge's Role in Plea Bargaining (pt. 1), 76 Colum. L. Rev. 1059, 1063-67, 1117 (1976) [hereinafter Alschuler, The Trial Judge's Role]; Alschuler, The Defense Attorney's Role in Plea Bargaining, supra note , at 1180, 1241, 1307-13; Albert W. Alschuler, The Prosecutor's Role in Plea Bargaining, 36 U. Chi. L. Rev. 50, 52-53, 85-105 (1968) [hereinafter Alschuler, The Prosecutor's Role]; Stephen J. Schulhofer, Plea Bargaining as Disaster, 101 Yale L.J. 1979, 1987-91 (1992); Stephen J. Schulhofer, Criminal Justice Discretion as a Regulatory System, 17 J. Legal Stud. 43, 52-60 (1988); Stephen J. Schulhofer, A Wake-Up Call from the Plea Bargaining Trenches, 19 Law & Soc. Inquiry 135, 136-38 (1994); accord Donald G. Gifford, Meaningful Reform of Plea Bargaining: The Control of Prosecutorial Discretion, 1983 U. Ill. L. Rev. 37, 45-65; John H. Langbein, Torture and Plea Bargaining, 46 U. Chi. L. Rev. 3, 12-19 (1978); David Lynch, The Impropriety of Plea Agreements: A Tale of Two Counties 19 Law & Soc. Inquiry 115, 117-26 (1994); Note, Plea Bargaining and the Transformation of the Criminal Process, 90 Harv. L. Rev. 564, 572-82 (1977).

23 Alschuler, Implementing the Criminal Defendant's Right to Trial, supra note , at 932-33; Alschuler, The Changing Plea Bargaining Debate, supra note , at 653, 657-58, 670-80, 711-13, 719-20; Alschuler, The Defense Attorney's Role, supra note , at 1197 n.55; Alschuler, The Prosecutor's Role, supra note , at 72-75, 82-83; Schulhofer, Plea Bargaining as Disaster, supra note , at 1983-86; Schulhofer, Criminal Justice Discretion as a Regulatory System, supra note , at 70-74; accord Gifford, supra note , at 45-65; Langbein, supra note , at 12-19; Note, Plea Bargaining and the Transformation of the Criminal Process, supra note , at 572-82; see also Alschuler, The Changing Plea Bargaining Debate, supra note , at 664-79 (criticizing plea bargaining because it puts a price on priceless rights and penalizes defendants who reject plea offers out of undue optimism or poor judgment); Michael O. Finkelstein, A Statistical Analysis of Guilty Plea Practices in the Federal Courts, 89 Harv. L. Rev. 293, 295, 304, 307, 311-12 (1975) (statistical analysis finding that prosecutors use deep plea discounts to induce defendats to plead guilty in cases in which they would be unable to secure convictions at trial because the evidence is too weak).

24 The one exception is Alschuler, The Changing Plea Bargaining Debate, supra note , at 657-82, 718-19 (arguing that plea bargains produce systematically unjust sentences that hinge on tactical decisions, do not reflect remorse or promote rehabilitation, and encourage defendants to think they have bought and sold priceless human liberty, penological objectives, and the right to be heard). The other, passing references to substantive values are in Santobello v. New York, 404 U.S. 257, 261 (1971) (s