The Real Problem with Plea Bargaining

Response - Online Edition - Volume 101

Most criminal law professors hate plea bargaining. While the precise reasons for their dislike may differ, they are nearly united in their condemnation of the practice. In his article, Plea Bargaining’s Uncertainty Problem, Professor Jeffrey Bellin takes aim at several of the common criticisms of plea bargaining, seeking to isolate the problem or problems that are uniquely caused by plea bargaining in order to offer reform proposals targeted at those particular problems.

I applaud Professor Bellin’s decision to look behind the conventional academic wisdom about plea bargaining. When our profession unites behind a particular position, it is important to subject that position to increased scrutiny and to guard against overly simplified arguments and overheated rhetoric. In his previous work, Professor Bellin has exposed shortcomings with other conventional wisdoms espoused by the criminal law academy.[2]

In his plea-bargaining article, Professor Bellin highlights an important shortcoming in the plea-bargaining literature—namely, that those who seek to abolish plea bargaining often fail to grapple with the fact that abolition would make the system much harsher than it currently is.[3] He also astutely observes that the election of progressive prosecutors may result in more defendants forgoing their right to trial because the plea bargains being offered are more favorable.[4]

Ultimately, however, I believe that Professor Bellin misses the mark in his criticisms of the academy’s dislike of plea bargaining. In framing his inquiry about the harms uniquely caused by plea bargaining, Professor Bellin both artificially limits the reasons for which plea bargaining should be disfavored, as well as the possible avenues of reform. And even within the framework that he chooses, I think that Professor Bellin has failed to identify the most important distinct harm that plea bargaining causes. Finally, I believe that the unique harm he identifies is not, in fact, caused by plea bargaining, and the particular plea-bargaining reforms that he proposes are unlikely to succeed.

I. Theoretical Framework for Problems and Solutions

Let me begin with a major theoretical premise underlying the article: That only by identifying plea bargaining’s “unique, problematic contribution to American criminal justice” can we correctly identify the reforms that are necessary to “fix plea bargaining.”[5] According to Professor Bellin, problems that are related to plea bargaining, but not caused by plea bargaining—such as harsh sentencing and the conviction of innocent defendants—should, be addressed directly through efforts “to mitigate unwarranted severity and reduce trial inaccuracy” rather than through reform of the plea bargaining system.[6] In other words, if plea bargaining did not solely cause a problem, then we should not reform plea bargaining to fix it.

As a matter of logic, I do not think this theoretical premise is correct. In a world of complex and overlapping problems, our ability to improve policies would be severely hampered by a rule that we change a policy only to address problems that are uniquely caused by that particular policy. Imagine, for example, a school district that is looking to change its curriculum in order to address low reading test scores. Using Professor Bellin’s framework, school officials should change their reading curriculum only to address literacy problems that had been caused by the existing curriculum, and not to address learning issues caused by other district policies, such as length of the school year or class size. Nor should they change their curriculum to address issues caused by external forces, such as poverty, parental illiteracy, or student learning disabilities. But taking that approach to curricular reform will likely set the school district up for failure. Reading scores are likely to improve only if school officials acknowledge that their reading curriculum does not exist in a vacuum but instead must be responsive to other policies and external factors that can affect literacy.

The same is true for criminal justice policy. The criminal justice system is a complex series of institutions, policies, and practices. Changes to one feature of the system can often affect other features. More specifically, changes external to the plea-bargaining process can affect plea bargains, and plea bargains can affect other aspects of the broader criminal justice system. For example, a robust bench trial system appears to reduce the rate of plea bargaining.[7] And plea bargaining appears to insulate unconstitutional police actions from judicial review.[8] As these examples illustrate, non-plea-bargaining policy changes could reform plea bargaining, and plea-bargaining reform could affect police policy. Thus, in assuming that only problems that are uniquely caused by plea bargaining should be the focus of plea-bargaining reform, Professor Bellin has artificially limited both the terms of debate and the likelihood of effective reform.

Professor Bellin’s theoretical premise also assumes that other problems with the criminal justice system can be effectively addressed on their own terms, without acknowledging the extent to which the prevalence of plea bargaining has likely hampered efforts at direct reform. Take, for example, the problem of harsh punishments. Those punishments are enshrined in statutes, which are enacted by legislatures. But legislatures assume prosecutors will not fully enforce statutes as written.[9] One of the major tools that prosecutors use in underenforcing statutes is their power to plea bargain. Perhaps legislatures would experience political pressure to revise statutes and make sentences less harsh if prosecutors maximally enforced statutes.[10] But so long as prosecutors continue to plea bargain and underenforce the law in most cases, such pressure is less likely to develop.

Finally, by artificially narrowing his inquiry about plea-bargaining reform to the problems that are uniquely caused by plea bargaining itself, Professor Bellin appears to undervalue the extent to which plea bargaining exacerbates various existing problems within the criminal justice system. For example, Professor Bellin dismisses conventional arguments about harsh post-trial sentences as a “product of baseline American criminal law enforcement.”[11] But plea bargaining does not merely “introduce[] an alternative to the non-negotiated baseline;”[12] it also contributes to that baseline problem: Lawmakers will enact laws with harsh punishments or refuse to revise laws to reduce existing punishments because those punishments are mitigated through plea bargaining.[13] Prosecutors also openly lobby legislatures to increase or maintain harsh sentences in order to give them leverage in plea negotiations.[14]

Professor Bellin discounts specific examples provided by plea bargaining’s critics of lawmakers and prosecutors acknowledging the role that plea bargaining plays in setting harsh baseline sentences. He characterizes these examples as demonstrating only that legislators have adopted harsher penalties to ensure cooperation, not to ensure plea bargains.[15] And because, Professor Bellin argues, prosecutors can always choose not to charge those who cooperate, legislatures “would still enact severe laws” even in the absence of plea bargaining.[16]

As an initial matter, it strikes me as misguided to assume that references to cooperation in legislative debates should be interpreted as anything other than implicit references to plea bargaining. After all, the cooperation that is incentivized by lengthy sentences is merely one type of plea bargain—a bargain in which the defendant agrees not only to plead guilty but also to provide evidence or testimony against others. In other words, these references to cooperation in legislative debates are direct evidence that lengthy sentences have been enacted or preserved in order to ensure one type of plea bargain.

Professor Bellin argues that lengthy sentences could still incentivize cooperation for defendants who have committed crimes in the absence of plea bargaining because prosecutors could simply use their charging power to decline to bring any charges against factually guilty defendants who cooperate.[17] He is certainly correct that prosecutors have the power to decline to bring charges against guilty defendants. But that does not suggest that these references to cooperation in legislative debates were divorced from plea bargaining or that they do not provide evidence that plea bargaining has exacerbated the problem of lengthy sentences. Many of the specific legislative debates that Professor Bellin dismisses contain references to safety valve sentencing provisions, making clear that legislators and those lobbying the legislature were discussing cooperation that included a plea bargain and a conviction, rather than cooperation that resulted in immunity for the defendant.[18]

In any event, to the extent that Professor Bellin is correct in assuming that cooperation should be viewed as distinct and separate from plea bargaining, there are examples of plea bargaining—and not cooperation—shaping legislative debates and decision-making about criminal codes. For example, Richard Denzer, a former prosecutor who served as the executive director of the major 1961 revision of New York’s criminal code, gave an interview about the process behind those revisions. When asked whether “the prevalence of guilty plea bargaining ha[d] a significant bearing on [his] drafting of any sections of the Penal Law,” Denzer responded: “Decidedly! We were very conscious of the negotiation process, and that’s the reason for our extensive degree structure. In criminal prosecutions, especially in New York County[,] it’s very important that one can take the negotiating process right down the line through the degrees.”[19]

More recently, former Attorney General William Barr wrote a joint letter to Congress opposing a legislative decrease to mandatory minimum sentences, stating: “Existing law already provides escape hatches for deserving defendants facing a mandatory minimum sentence. Often, they can plea bargain their way to a lesser charge; such bargaining is overwhelmingly the way federal cases are resolved.”[20] In a similar vein, Senator Tom Cotton publicly opposed criminal justice reforms that would allow the release of defendants convicted for certain low-level crimes arguing that “to the extent people are in federal prison for low-level convictions, they typically pled down from more serious charges.”[21] In other words, as these examples show, the general practice of plea bargaining—and not simply plea bargains that involve cooperation agreements—shapes the decisions and the arguments of legislators and those who lobby them.

More fundamentally, it is hard to know how to address Professor Bellin’s hypotheses about legislative incentives. He may be correct that, even in the absence of plea bargaining, legislatures still would have incentives to enact harsh sentences with the understanding that they would apply only to those who do not cooperate and thus avoid conviction.[22] He is almost certainly correct that plea bargaining is just one of many factors that affect legislative decisions.[23] But it is also quite clear that legislators and many of those who lobby them account for plea bargaining when making decisions about criminal sentences, and they see the existence of plea bargaining for both cooperators and noncooperators as a reason to keep sentences harsh. While we cannot know how a system without plea bargaining would decide legislative sentencing questions, we do know that plea bargaining has shaped the current and past legislative debates. Presenting conclusive evidence resolving a counterfactual is not necessary to confidently say that plea bargaining has, at a minimum, exacerbated the problem of harsh sentences in this country.

There is also reason to believe that plea bargaining exacerbates the problem of innocent defendants pleading guilty. The power that plea bargaining gives to prosecutors allows them to offer plea deals that no rational defendant would turn down. A rational defendant would accept any plea bargain in which her expected punishment is less than the punishment she will receive after trial. So, for example, an innocent defendant who is facing a sentence of five years if convicted at trial should accept a plea offer of less than six months if she believes there is a 10% chance she will be convicted at trial.[24] And an innocent defendant who is facing any jail time at all if convicted should rationally accept an offer of diversion or probation. That plea bargaining leads to innocent people pleading guilty is not merely an abstract concept. There is ample evidence suggesting that a nontrivial number of innocent people plead guilty.[25]

Professor Bellin acknowledges that innocent people sometimes plead guilty and that a large difference between the sentence in a plea agreement and the expected post-trial sentence can create so much pressure “that it can become rational for even the innocent to plead guilty.”[26] And while Professor Bellin concedes that the conviction of innocent defendants represents a grave injustice, he argues that “plea bargaining contributes to this injustice only if the defendant would have been acquitted at trial (or the case would have been dismissed).”[27] If the innocent defendant would have been convicted at trial, then plea bargaining “mitigates the injustice” because the defendant will serve a shorter sentence than if he had proceeded to trial.[28]

Whether plea bargaining results in the conviction of more innocent defendants than do criminal trials is an empirical question. Professor Bellin appears to assume that trials generate more erroneous convictions: He notes that, in Judge Jed Rakoff’s criticism of plea bargaining, Judge Rakoff cites studies establishing that 10% of the wrongfully convicted pleaded guilty, and thus, by implication, the other 90% were convicted at trial.[29] Professor Bellin concludes that these figures fail to establish that plea bargaining “causes (or aggravates) . . . punishment of the innocent.”[30]

It is entirely fair to criticize Judge Rakoff’s reliance on particular studies to prove that plea bargaining, as opposed to trials, are a significant cause of wrongful convictions. But it is a mistake to conclude that those studies accurately reflect the contribution that plea bargaining makes to the problem of punishing the innocent. To the contrary, there are reasons to believe that these studies of the exonerated substantially undercount the number of innocent defendants who pleaded guilty.

As a legal matter, it is more difficult for defendants who plea bargain to later obtain exoneration. In my home state of North Carolina, for example, the Innocence Inquiry Commission, which is tasked with investigating and screening claims of factual innocence, requires people who plead guilty to convince every person on the eight-member Commission that they are probably innocent before they get a chance to prove their innocence to the panel of judges.[31] Those who were convicted after trial need only convince a majority of the members.[32] Nationwide, approximately fifteen states exclude defendants who pleaded guilty from their DNA-testing statutes.[33] And across the country, innocent defendants who pleaded guilty are often unable to obtain exoneration because their plea agreements included a waiver of the right to appeal and the right to collateral review.[34]

As a practical matter, innocent defendants who plead guilty are less likely to be exonerated than innocent people who insist on going to trial. That is because the two groups are not similarly situated. For example, the limited social science evidence on plea bargaining indicates that defendants who are facing more serious charges are less likely to plead guilty and more likely to insist on a trial.[35] This difference between the two groups—that defendants who proceed to trial were more likely to have been facing serious charges than those who pleaded guilty—could skew exoneration rates in at least two ways.

First, as a group, innocent defendants who went to trial are likely to spend more time in prison than innocent defendants who pleaded guilty—that is both because people facing more serious charges are more likely to go to trial and also because those who plead guilty receive shorter sentences than those who go to trial. Longer sentences give defendants more of an opportunity to exonerate themselves with new evidence that was not introduced at trial. Most courts will not allow a defendant to introduce new evidence, including new evidence of their innocence, on direct appeal; as a consequence, defendants must wait until collateral review has begun to litigate those issues which could lead to their exoneration.[36] As Professor Eve Brensike Primus has explained:

[M]ost defendants have served their full sentences by the time they reach the collateral review stage. Under the current system, only defendants sentenced to more than four or five years in prison have an incentive to challenge their convictions on collateral review, because it takes that long to exhaust the appellate process in many jurisdictions.[37]

Because they serve longer sentences, innocent defendants who were convicted after trials will be overrepresented in the group of prisoners who are still incarcerated after four or five years as compared to innocent defendants who pleaded guilty. Consequently, we should expect innocent defendants who went to trial to make up more than their fair share of innocent defendants who are ultimately exonerated. Conversely, we should expect innocent people who pleaded guilty to make up a disproportionate percentage of the innocent who are never exonerated.

Further skewing the pool of innocent defendants who are ultimately exonerated is the gold standard for post-conviction exonerations—DNA evidence. DNA-based exonerations are not possible in most cases because those cases do not involve DNA evidence.[38] DNA evidence is largely confined to cases involving homicide and sexual assault. These are very serious charges, and because defendants facing more serious charges are more likely to go to trial than plead guilty,[39] we should expect DNA exonerees to have proceeded to trial at a higher rate than the pool of all innocent defendants.

To be clear, I do not know whether more innocent defendants plead guilty than insist on a trial. Nor do I know whether those innocent defendants who pleaded guilty would have been convicted after trial. However, Professor Bellin does not know the answers to these questions either. The limited empirical data that he points to in suggesting that plea bargaining does not cause or aggravate punishment of the innocent is not only clearly incomplete, but it also fails to account for numerous legal and practical issues that lead to the disproportionate exoneration of innocent defendants who proceeded to trial rather than pleading guilty. And while some steps have been taken in recent decades to reduce the number of innocent defendants who are convicted at trial,[40] the same is not true for defendants who plead guilty. To the contrary, legislatures have increased the scope of criminal law and increased punishments more than they have decreased them,[41] which increases prosecutors’ plea-bargaining leverage and offers them even greater opportunity to offer plea bargains that no rational defendant would refuse.

II. Plea Bargaining’s Fundamental Problem

In my view, the major problem of plea bargaining is that it has fundamentally warped the criminal justice system by prioritizing the efficient disposition of cases above constitutional rights and the rule of law. Plea bargaining removes all of the procedural protections associated with a criminal trial, and it fails to replace those protections with any other formal process or rights. As Professor Darryl Brown puts it, in modern American plea bargaining, “the law regulates as little as possible.”[42]

In order to conserve resources, plea bargains rely on negotiations between the parties, rather than trials before neutral juries and judges to resolve questions of guilt. But the courts will generally refuse to police the content of those negotiations.[43] Prosecutors are free to threaten defendants with rarely imposed sentencing enhancements,[44] charges against their friends and families, and even the death penalty[45] in order to pressure defendants into pleading guilty.[46] So long as there is a legal basis for those threats, they are legally permitted. What is more, prosecutors can demand the defendant waive more than just his right to a trial in return for the plea agreement. Prosecutors can (and have) required defendants to waive both pretrial rights (such as the right to discovery and the right to file suppression motions) and post-trial rights (such as the right to appeal, the right to postconviction review, and the possibility of compassionate release from prison) as part of a plea bargain.[47]

In refusing to place limits on how much leverage prosecutors can wield during plea negotiations and what rights they can require the defendant to waive, courts have emphasized the efficiency benefits of plea bargains. In the case that first recognized the constitutionality of plea bargaining, the Supreme Court noted that, without plea bargaining, “the States and the Federal Government would need to multiply by many times the number of judges and court facilities.”[48] Chief Justice Warren Burger, who wrote that opinion, gave a public address in which he stated that the United States absolutely depends on the vast majority of defendants pleading guilty, and it could not afford to have even 20% of cases go to trial.[49] In more recent years, the Court justified its refusal to require prosecutors to turn over discovery during plea negotiations, in part, on the grounds that requiring the government to disclose such information would make the plea bargaining less efficient and thus “could lead the Government . . . to abandon its heavy reliance upon plea bargaining in a vast number . . . of federal criminal cases”—an outcome that the Justices appeared eager to avoid.[50]

The protections that plea bargaining removes are not merely of abstract value; some of them are designed to ensure the accuracy of convictions. Removing the ability of a defendant to demand discovery during plea negotiations, for example, distorts the negotiations between defense attorneys and prosecutors. Defense attorneys do not have access to the information that allows them to accurately estimate the chances of conviction at trial—including exculpatory evidence that could indicate the defendant is factually innocent.[51] And because prosecutors need not turn over discovery, they have less incentive to review that discovery themselves.[52] Waiving postconviction rights also reduces the accuracy of convictions. Appeals allow for the correction of any errors on the part of the trial court.[53] And collateral attacks permit the consideration of newly discovered evidence that can cast doubt on a defendant’s guilt.[54]

Even the waiver of trial itself—the bedrock of plea bargaining—likely reduces the accuracy of convictions. A trial allows a neutral third party to assess the factual evidence to determine whether a defendant is guilty of the crime charged. In contrast, plea bargaining leaves the determination of guilt in the hands of prosecutors, whose decisions are likely affected by cognitive bias.[55] In addition, when the question of factual guilt is taken away from jurors, it is no longer treated as a binary question to be resolved either in the affirmative or the negative; instead factual guilt becomes a matter of negotiation for lawyers.[56] The result is a conviction that everyone knows does not definitely resolve the factual disputes between the parties but instead represents a compromise. Consequently, convictions are no longer regarded as an accurate record of what the defendant actually did.[57]

This criticism of plea bargaining—that it has removed the procedural protections of trial and failed to provide any new protections—is hardly new. It is, for example, incredibly well-developed by Professor Darryl Brown in his 2016 book, Free Market Criminal Justice: How Democracy and Laissez Faire Undermine the Rule of Law.[58] A form of the argument also appears in at least one of the articles that Professor Bellin discusses.[59] And although this criticism satisfies Professor Bellin’s theoretical premise—it is a problem that is caused by plea bargaining itself—he does not address the criticism in any detail or identify it as the basis for plea-bargaining reform.

III. Uncertainty and Plea Bargaining

The major, distinct problem caused by plea bargaining that Professor Bellin identifies is uncertainty—namely, the inability of a defendant to assess whether the plea agreement being offered is better or worse than the likely outcome at trial. After identifying this problem, Professor Bellin then seeks to reform plea bargaining by focusing on the uncertainty that plea bargaining causes. I see two flaws with this analysis.

First, uncertainty about trial outcomes is not a problem that is unique to plea bargaining, nor is it caused by the practice of plea bargaining. A defendant who is deciding to proceed to trial or to plead guilty without a plea agreement (sometimes referred to as an “open plea”) also faces uncertainty. She will not know what sentence she is facing if convicted at trial, nor will she know what discount, if any, she will receive from the trial judge if she pleads guilty.[60] In addition, she does not know how quickly she will need to plead guilty in order to receive that sentencing discount: Can she wait until the eve of trial, or must she plead soon after arraignment in order for the trial court to show mercy?

Certainty about the outcome after trial may be of greater value to the defendant in plea bargaining because certainty could operate as a bargaining chip in the defense attorney’s negotiation with the prosecutor. Especially when the difference between the plea offer and the post-trial sentence is not large, the defense attorney could use that information to convince the prosecutor to increase the discount being offered in order to make the plea bargain more attractive.[61] But certainty might also lead the prosecutor to offer a less favorable plea deal to the defendant—once it is clear to the prosecutor what the post-trial sentence would be, she might be unwilling to offer significantly less than what she could obtain after trial. Put differently, certainty may give the defendant more information about the difference between the plea agreement and the potential outcome, but it is not clear what effect certainty might have on the plea negotiations themselves.[62]

In any event, plea bargaining is not causing the problem of uncertainty; at most plea bargaining exacerbates the consequences of the problem because defendants must decide not only whether to go to trial or to plead open but also whether to accept a plea bargain. What is more, increasing the certainty of trial outcomes is something that can be addressed on its own rather than through plea-bargaining reform.[63] Consequently, uncertainty is no different than other problems, such as sentencing severity, that Professor Bellin dismisses earlier in the article.[64]

Second, and more important, Professor Bellin does not grapple with the fact that the very nature of plea bargaining—that it gives prosecutors leverage to force defendants to waive their rights—could frustrate the proposals that he offers. Professor Bellin proposes that, in order to reduce uncertainty, defendants should be given access to discovery before plea bargaining, and defense lawyers should conduct more vigorous investigations to uncover evidence that the government may not possess.[65] Neither proposal is compatible with the fundamental problem of modern American plea bargaining that I discuss above—namely that plea bargaining leaves everything up for negotiation. Prosecutors could condition their plea bargaining offers on a defendant’s waiver of the right to discovery. They can also condition plea bargaining on a waiver of the effective right to counsel, if not an absolute waiver on the right to counsel all together.

My claim that prosecutors could use plea bargaining to circumvent Professor Bellin’s proposals is based on more than mere conjecture. There are examples where it has already happened.

Arizona provides one stark example. That state has a robust criminal discovery rule, which requires prosecutors turn over discovery no more than thirty days after arraignment.[66] The prosecutor’s office in Phoenix has adopted a policy that circumvents that state discovery law.[67] In order to avoid discovery obligations, the prosecutor’s office demands pre-arraignment guilty pleas from defendants. Defendants and their lawyers have been informed that the plea agreement offer will expire at the time of indictment. After they are indicted and arraigned, defendants will be entitled to discovery, but they will either receive a less favorable plea offer or no offer at all.[68] This policy is consistent with Supreme Court precedent,[69] and attempts to challenge the policy on constitutional grounds have failed.[70]

Robust defense attorney investigations can also fall victim to prosecutors’ near-exclusive power over plea bargaining. Prosecutors can (and do) place time limits on plea offers that can prevent defense attorneys from conducting in-depth investigations.[71] Some prosecutors even require defendants to waive their right to effective counsel as part of the plea negotiation.[72] Such waivers would insulate any investigative failures by defense attorneys from appellate or collateral review. Indeed, for a while, such waivers were common in federal cases. The Department of Justice eventually issued a memo stating that federal prosecutors should stop seeking such waivers.[73] But the memo insisted that the practice was entirely constitutional, and there is little reason to think that the U.S. Supreme Court would disagree with that assessment.

Because the content of plea negotiations is essentially exempt from judicial review, prosecutors could insist that defendants waive their right to counsel altogether in order to receive a favorable plea deal. There is evidence of at least one state prosecutor doing precisely that—offering favorable plea deals only if defendants do not ask for the appointment of counsel.[74] While such a plea negotiation tactic appears to violate professional ethics rules,[75] it is not obviously unconstitutional under prevailing Supreme Court doctrine.

To be clear, I agree with Professor Bellin that defendants should have more information when deciding whether to plead guilty. Indeed, the theoretical foundation of plea bargaining—that it allows the parties to bargain in the shadow of trial—depends on defendants having a clear sense of the likely trial outcome. But in refusing to regulate the process of plea bargaining and instead relying on what Professor Darryl Brown has characterized as a “free market” approach to criminal justice,[76] the Supreme Court has made “fixing” plea bargaining an exceptionally difficult task. In order for plea-bargaining reforms to work, there must be some limit on the ability of prosecutors to control the substance of plea negotiations. Otherwise, defendants can be required to waive any benefits of reform as a condition of a plea agreement.[77]

Unless and until this fundamental issue with plea bargaining is addressed—namely, that it gives enormous leverage to prosecutors to extract guilty pleas and force defendants to waive even nontrial rights—then indirect plea-bargaining reforms may be the best approach.[78] For example, reducing harsh punishments and improving the accuracy of trial outcomes could indirectly reform plea bargaining because they would reduce prosecutors’ leverage and make trial a more attractive alternative to pleading guilty. These second-best solutions do not meet Professor Bellin’s theoretical framework for reforming plea bargaining. But they avoid plea bargaining’s biggest problem.


Professor Bellin is correct that plea bargains are the best option for most defendants—plea bargaining allows them to negotiate an outcome that is less harsh than the likely outcome after a trial. But focusing only on favorable outcomes for individual defendants can obscure the extent to which plea bargaining has fundamentally warped American criminal justice. Whatever their shortcomings, criminal trials provide a clear opportunity to resolve factual disputes, they involve multiple institutions in that resolution, and both their procedures and their outcomes are subject to public scrutiny. Plea bargains, in contrast, obscure the facts surrounding crimes, concentrate decision-making in the hands of prosecutors, and fail to provide a clear or public procedural framework for those decisions. In this regard, plea bargaining, at least as it currently exists in America, represents a dramatic departure from the constitutional values that are supposed to animate our criminal justice system.[79] Professor Bellin is correct that plea bargaining may lead to better outcomes in some individual cases, but on balance it has made our system worse.[80]

More fundamentally, it is not clear that delivering better outcomes to individual defendants should be the overarching goal of plea-bargaining reform. Favorable outcomes are not always accurate or just outcomes. Indeed, one of the oldest objections to plea bargaining is that it results in outcomes that are too favorable to defendants.[81] Plea bargaining empowers prosecutors to dictate outcomes, and they can use that power to offer overly lenient pleas. Professor Bellin tells us that “the normative baseline—the ‘punishment the defendant should receive’—is contested,”[82] and so he may reject the idea that a plea bargain can be too lenient. But there are examples, such as the plea bargain that was given to Jeffrey Epstein,[83] that result in a consensus about punishment; the public may not agree precisely how much punishment Epstein should have received, but they agree he should have received more.

In addition, plea bargaining may prevent a public consensus about punishment from developing. Plea bargains are often negotiated in secret, leading to less public scrutiny and debate about appropriate punishment levels. And when prosecutors are challenged about their plea bargains, they often justify them on the grounds of what evidence or resources are available[84] rather than on terms that would facilitate a public discussion about normative baselines. These are yet more problems that plea bargaining causes and possible areas for reform.

It is because plea bargaining has fundamentally changed the criminal justice system that it is so widely disliked. That there are so many different criticisms of plea bargaining is not necessarily evidence that the criticism is “muddled,”[85] but rather that plea bargaining has caused or exacerbated so many different problems. Seeking to find a discrete and exclusive problem that plea bargaining causes can blind us to the far-reaching consequences of having replaced our system of criminal trials with a system of plea bargains[86] and then having failed to regulate plea bargains in any meaningful way. Providing defendants more certainty about post-trial outcomes could assist them in making better plea-bargaining decisions, but it does not “fix” the problems that plea bargaining causes.

  1. † Ransdell Distinguished Professor of Law and Director of the Prosecutors and Politics Project, University of North Carolina School of Law.
  2. .See, e.g., Jeffrey Bellin, The Power of Prosecutors, 94 N.Y.U. L. Rev. 171, 174 (2019).
  3. .See, e.g., Jeffrey Bellin, Plea Bargaining’s Uncertainty Problem, 101 Texas L. Rev. 539, 543 (2023).
  4. .Id. at 539.
  5. .Id. at 547 (emphasis omitted).
  6. .Id.
  7. .See generally Stephen J. Schulhofer, Is Plea Bargaining Inevitable?, 97 Harv. L. Rev. 1037 (1984).
  8. .See Carissa Byrne Hessick, Punishment Without Trial: Why Plea Bargaining is a Bad Deal 171–75 (2021).
  9. .For a theoretical account of this relationship between legislatures and prosecutors, see William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 511, 546–48 (2001). For real-world examples of lawmakers criticizing prosecutors for enforcing the law as written rather than exercising “better judgment,” see Carissa Byrne Hessick & Joseph E. Kennedy, Criminal Clear Statement Rules, 97 Wash. U. L. Rev. 351, 354 n.7 (2019) (quoting Matt Bokor, Prosecutors Have Little Sympathy for Senior Gamblers, Associated Press, Feb. 4, 1982, Lexis).
  10. .This is the premise of Daniel Epps, Adversarial Asymmetry in the Criminal Process, 91 N.Y.U. L. Rev. 762 (2016).
  11. .See Bellin, supra note 2, at 551.
  12. .See id. at 552.
  13. .See, e.g., Russell M. Gold, Carissa Byrne Hessick & F. Andrew Hessick, Civilizing Criminal Settlements, 97 B.U. L. Rev. 1607, 1617–18 (2017).
  14. .Rachel E. Barkow, Administering Crime, 52 UCLA L. Rev. 715, 728 n.25 (2005) (citing examples from the congressional record “of the Department of Justice requesting more stringent sentencing laws because it makes prosecutors’ jobs easier”).
  15. .Bellin, supra note 2, at 561 (distinguishing cooperation from guilty pleas and characterizing the desire to incentivize cooperation as “a related but distinct goal” of plea bargaining).
  16. .Id.
  17. .See id. (“[D]efendants who cooperate do not always plead guilty. Some avoid charges altogether.”).
  18. .For example, the statement in opposition to reducing mandatory minimum sentences by Senator Chuck Grassley, which Professor Bellin references on p. 561 n.107, argues that it is not necessary to reduce mandatory minimum sentences because so few drug couriers actually serve such sentences; the rest end up receiving lower sentences because of statutory sentencing safety valves or because they cooperated with prosecutors. See Bellin, supra note 2, at 561 n.107 (citing Russell M. Gold, Carissa Byrne Hassock & F. Andrew Hassock, Civilizing Criminal Settlements, 97 B. U. L. Rev. 1607, 1618 n.32 (2017) (quoting 161 Cong Rec. S955-02, S963 (daily ed. Feb. 12, 2015) (statement of Sen. Grassley))). Senator Grassley identified the average sentence for drug couriers, which was below the applicable mandatory minimum but still above three years, and he remarked “[t]hat seems to be an appropriate level.” 161 Cong Rec. 2240 (2015) (statement of Sen. Grassley).
  19. .Drafting a New Penal Law for New York: An Interview with Richard Denzer, 18 Buff. L. Rev. 251, 258 (1969) (footnote omitted). Denzer’s statement about the influence of plea bargaining is quoted in two seminal articles on modern plea bargaining. See Albert W. Alschuler, The Trial Judge’s Role in Plea Bargaining, Part I, 76 Colum. L. Rev. 1059, 1145 & n.273 (1976); George Fisher, Plea Bargaining’s Triumph, 109 Yale L.J. 857, 1033 n.677 (2000).
  20. .Letter from William P. Barr, Former U.S. Att’y Gen., et al. to Harry Reid, Majority Leader, U.S. Senate, and Mitch McConnell, Minority Leader, U.S. Senate (May 12, 2014), https://‌‌/site‌/index.php‌/resources‌/letters‌/57-may-2014-letter-opposing-mandatory-minimums-reduction‌/file [https://‌‌/NH4F-DR8K].
  21. .Sen. Tom Cotton, What’s Really in Congress’s Justice-Reform Bill, Nat’l Rev. (Nov. 26, 2018, 10:48 AM) https://‌‌/2018‌/11‌/first-step-criminal-justice-reform-bill-whats-in-it [].
  22. .Bellin, supra note 2, at 561 (“[E]ven if formal plea bargains were forbidden, a legislature trying to incentivize cooperation would still enact severe laws and intend that they apply in some cases—specifically, to defendants who do not cooperate.”).
  23. .See id. at 561–64.
  24. .For an expanded discussion about expected punishments and how they affect plea bargaining, see Hessick, supra note 7, at 37–38.
  25. .See, e.g., Ram Subramanian, Léon Digard, Melvin Washington II & Stephanie Sorage, Vera Inst. for Just., In the Shadows: A Review of the Research on Plea Bargaining 44–45 & 64 nn.203–07 (2020), https://‌‌/downloads‌/publications‌/in-the-shadows-plea-bargaining.pdf [] (describing and collecting sources that rely on self-reporting); Ronald F. Wright, Trial Distortion and the End of Innocence in Federal Criminal Justice, 154 U. Pa. L. Rev. 79, 84–86 (2005) (employing a statistical analysis of the decreased acquittal rate to support a hypothesis that defendants who previously would have been acquitted at trial are pleading guilty).
  26. .Bellin, supra note 2, at 567.
  27. .Id. at 568.
  28. .Id.
  29. .Id. at 567–68 (citing Jed S. Rakoff, Why Innocent People Plead Guilty, N.Y. Rev. Books (Nov. 20, 2014), []).
  30. .Id. at 568.
  31. .See N.C. Gen. Stat. §§ 15A-1462(a), -1467(a), -1468(c) (2006).
  32. .See id. § 15A-1468(c).
  33. .Samuel R. Wiseman, Waiving Innocence, 96 Minn. L. Rev. 952, 955 n.14 (2012).
  34. .Nancy J. King & Michael E. O’Neill, Appeal Waivers and the Future of Sentencing Policy, 55 Duke L.J. 209, 212 (2005) (finding that nearly two-thirds of a sample of federal plea agreements included an appeal waiver); Susan R. Klein, Aleza S. Remis & Donna Lee Elm, Waiving the Criminal Justice System: An Empirical and Constitutional Analysis, 52 Am. Crim. L. Rev. 73, 87 (2015) (finding that more than two-thirds of federal boilerplate plea agreements include waivers of the right to collaterally attack the conviction).
  35. .See, e.g., Celesta A. Albonetti, Race and the Probability of Pleading Guilty, 6 J. Quantitative Criminology 315, 323 (1990); Jon’a Meyer & Tara Gray, Drunk Drivers in the Courts: Legal and Extra-Legal Factors Affecting Pleas and Sentences, 25 J. Crim. Just. 155, 157 (1997).
  36. .See Commonwealth v. Grant, 813 A.2d 726, 734–37 (Pa. 2002) (surveying state and federal court opinions on the matter), clarified on denial of reargument, 821 A.2d 1246 (Pa. 2003) (per curiam), and modified by Commonwealth v. Bethea, 828 A.2d 1066 (Pa. 2003), and abrogated on other grounds by Commonwealth v. Bradley, 261 A.3d 381 (Pa. 2021).
  37. .Eve Brensike Primus, Structural Reform in Criminal Defense: Relocating Ineffective Assistance of Counsel Claims, 92 Cornell L. Rev. 679, 693 (2007).
  38. .Brandon L. Garrett, Convicting the Innocent Redux, in Wrongful Convictions and the DNA Revolution: Twenty-Five Years of Freeing the Innocent 40, 54 (Daniel S. Medwed ed., 2017) (“[M]ost criminal cases have no DNA to test.”); Richard A. Leo, Has the Innocence Movement Become an Exoneration Movement? The Risks and Rewards of Redefining Innocence, in Wrongful Convictions and the DNA Revolution: Twenty-Five Years of Freeing the Innocent 57, 73 (Daniel S. Medwed ed., 2017) (“An estimated 80–90 percent of cases do not involve biological evidence . . . .” (citing Daniel S. Medwed, Beyond Biology: Wrongful Convictions in the Post‑DNA World, 2008 Utah L. Rev. 1, 1–2 nn.6–8 (2008))).
  39. .See Albonetti, supra note 34; Meyer & Gray, supra note 34.
  40. .Stephanie Roberts Hartung, Post-Conviction Procedure: The Next Frontier in InnocenceReform, in Wrongful Convictions and the DNA Revolution: Twenty-Five Years of Freeing the Innocent 247, 248–50 (Daniel S. Medwed ed., 2017).
  41. .See Carissa Byrne Hessick, Ronald F. Wright & Jessica Pishko, The Prosecutor Lobby, 79 Wash. & Lee L. Rev. (forthcoming 2023) (manuscript at 32 tbl.1) (reporting that, in the years 2015–2018, state legislatures passed nearly 600 bills to increase punishments and fewer than 300 bills to reduce punishment).
  42. .Darryl K. Brown, Free Market Criminal Justice: How Democracy and Laissez Faire Undermine the Rule of Law 97 (2016).
  43. .See id. at 94–95, 102–104 (recounting the relevant history of plea-bargaining doctrine).
  44. .See Bordenkircher v. Hayes, 434 U.S. 357, 358–59, 365 (1978).
  45. .See, e.g., People v. Granberry, 256 N.E.2d 830, 832 (Ill. 1970); see also Brady v. United States, 397 U.S. 742, 743, 758 (1970).
  46. .E.g., Miles v. Dorsey, 61 F.3d 1459, 1468–69 (10th Cir. 1995); United States v. Pollard, 959 F.2d 1011, 1020–21 (D.C. Cir. 1992).
  47. .See supra text accompanying note 32 and infra text accompanying notes 64–69.
  48. .Santabello v. New York, 404 U.S. 257, 260 (1970).
  49. .Burger, The State of the Judiciary—1970, 56 A.B.A. J. 929, 931 (1970).
  50. .United States v. Ruiz, 536 U.S. 622, 632 (2002).
  51. .E.g., Mansfield v. Williamson County., 30 F.4th 276, 281 (5th Cir. 2022) (Higgenbotham, J., concurring) (explaining that the Fifth Circuit “has consistently held” that the right to the disclosure of exculpatory evidence “focuses on the integrity of trials and does not reach pre-trial proceedings leading to guilty pleas” and thus the Fifth Circuit has twice explicitly “held that there is no constitutional right to exculpatory evidence during plea bargaining” (first citing Matthew v. Johnson, 201 F.3d 353, 361 (5th Cir. 2000); then citing United States v. Conroy, 567 F.3d 174, 178 (5th Cir. 2009) (per curiam); and then citing Alvarez v. City of Brownsville, 904 F.3d 382, 392 (5th Cir. 2018) (en banc))), cert. denied, 143 S. Ct. 486 (2022).
  52. .When conducting interviews for my recent book on plea bargaining, I often heard about how prosecutors failed to evaluate the evidence in a case before making a plea offer. See Hessick, supra note 7, at 58 (describing a defense attorney having to request that prosecutors evaluate available video surveillance evidence in order to alter a plea offer or dismiss charges); id. at 120–21 (describing a plea bargain on a trespassing charge for a defendant who was arrested at his own residence).
  53. .See United States v. Townsend, No. 19-20840, 2021 WL 777191, at *4 (E.D. Mich. Mar. 1, 2021) (rejecting a plea agreement with an appeal waiver, inter alia, because it would preclude the correction of any errors), mandamus granted by In re United States, 32 F.4th 584 (6th Cir. 2022).
  54. .See supra note 34 and accompanying text.
  55. .See Brown, supra note 41, at 169–70 (stating that “it matters a great deal who decides” which is “why the American system supposedly (formerly?) valued juries, and why Americans still think of prosecutors . . . as different from judges”); cf. Sandra Guerra Thompson, Cops in Lab Coats: Curbing Wrongful Convictions through Independent Forensic Laboratories 129–33, 181 (2015) (discussing how cognitive bias suggests that forensic scientists should be employed outside of the law enforcement bureaucracy); Rachel E. Barkow, Prosecutorial Administration: Prosecutor Bias and the Department of Justice, 99 Va. L. Rev. 271, 313, 341–42 (2013) (similar).
  56. .See, e.g., Hessick, supra note 7, at 30–31 (recounting public statements by an elected district attorney that juries should not resolve factual questions of self-defense or provocation).
  57. .See, e.g., Cotton, supra note 20; see also Thea Johnson, Fictional Pleas, 94 Ind. L.J. 855, 864–66 (2019) (providing numerous examples of plea bargains that result in convictions that are acknowledged to be fictional as a matter of fact or a matter of law).
  58. .See generally Brown, supra note 41. Although Professor Bellin cites some of Professor Brown’s other work, he does not discuss this book nor this criticism of plea bargaining. See Bellin, supra note 2, at 547 n.39, 549 nn.50–51, 555 n.80, 558–59 n.96.
  59. .For example, Gold, et al., supra note 12, criticizes criminal plea bargaining because, especially as compared to the civil settlement system, it is not governed by procedures or process. See id. at 1609–10. The article also discusses how prosecutorial leverage would need to be reduced in order to keep prosecutors from requiring defendants to waive any settlement-oriented procedures. See id. at 1611.
  60. .Professor Bellin acknowledges that judges impose trial penalties in the absence of plea bargains. Bellin, supra note 2, at 556, and he acknowledges both that “[i]n a jury system, some uncertainty is unavoidable,” id. at 547, and that post-trial sentences can be unpredictable because of judicial sentencing discretion. Id. at 565. But he insists that the “uncertainty that matters is uncertainty about whether the defendant should accept a plea offer.” Id. at 547. And yet, that uncertainty exists only because there is uncertainty about outcomes after trials—if the outcome after trial were known, then there would be no uncertainty about whether to accept the plea bargain. In other words, plea bargaining changes the consequences of the post-trial uncertainty by introducing another decision that the defendant must make. But that does not seem sufficient to satisfy Professor Bellin’s theoretical premise that “[i]f plea bargaining is not the source of these problems, it is rarely the right place to fix them.” Id. at 553.
  61. .That is because the larger the difference between the expected punishment and the prosecutor’s plea offer, the more attractive the offer becomes. See Hessick, supra note 7, at 37–38.
  62. .In this respect, certainty adds another ingredient to the “complex stew of incentives” that led Professor Bellin to discount other critiques of plea bargaining as promising grounds for reform. Bellin, supra note 2, at 559–64. To be sure, Professor Bellin convincingly demonstrates that more certainty can allow defendants to more accurately predict when plea bargaining is the less punitive option. See, e.g., id. at 573–74. But allowing defendants to make better choices does not necessarily leave defendants better off if the choices themselves become less favorable.
  63. .For example, the Federal Sentencing Guidelines made trial outcomes and guilty plea discounts far more certain. See U.S. Sentencing Comm’n, Fifteen Years of Guideline Sentencing: An Assessment of How Well the Federal Criminal Justice System is Achieving the Goals of Sentencing Reform xvi (2004). That certainty has decreased somewhat in the wake of United States v. Booker, 543 U.S. 220 (2005), which made the Guidelines advisory rather than mandatory. See id. at 245–46.
  64. .Bellin, supra note 2, at 545.
  65. .Id. at 575–76. Bellin also proposes that defendants be given more access to statistical evidence about trial outcomes in the relevant jurisdictions. Id. at 576. This proposal does not suffer from the defect of the other two—namely that defendants can be forced to bargain these rights away as part of the plea-bargaining process. However, I have reservations about how helpful such information would be in light of how few trials occur in some jurisdictions. See, e.g., Frank O. Bowman, III, American Buffalo: Vanishing Acquittals and the Gradual Extinction of the Federal Criminal Trial Lawyer, 156 U. Pa. L. Rev. PENNumbra 226, 237 (2007) (“In 2002, thirty-one out of the ninety-four federal districts saw fewer than ten trials. The two districts covering the state of Wisconsin boasted eleven trials between them. Vermont reported zero trials in 2002 and only two in 2001.” (footnotes omitted)). What is more, this proposal does not merely reform plea bargaining. It would also benefit defendants who are deciding whether to plead guilty without an offer from the prosecutor, and it has the benefit of making sentencing decisions more transparent, which could lead to sentencing reform.
  66. .See Ariz. R. Crim. P. 15.1.
  67. .See Hessick, supra note 7, at 55–57 (describing the policy).
  68. .Id.
  69. .See United States v. Ruiz, 536 U.S. 622, 628–29 (2002).
  70. .See, e.g., Luckey v. Mitchell, No. CV 21-01168, slip op. at 28 (D. Ariz. Sept. 17, 2022) (granting defendants’ motion to dismiss).
  71. .See, e.g., United States v. Pickering, 178 F.3d 1168, 1174 (11th Cir. 1999) (confirming a prosecutor’s power “to offer a defendant an ‘exploding’ plea bargain with a short fuse”).
  72. .See Klein et al., supra note 33, at 87 (reporting that a sizeable percentage of federal plea agreements in a study sample contained an effective assistance of counsel waiver).
  73. .See Memorandum from James M. Cole, Deputy Att’y Gen., U.S. Dep’t of Just., to All Federal Prosecutors, Department Policy on Waivers of Claims of Ineffective Assistance of Counsel (Oct. 14, 2014), [].
  74. .See Hessick, supra note 7, at 57–59.
  75. .See Model Rules of Pro. Conduct r. 3.8(c) (Am. Bar Ass’n 2020); see also Hessick, supra note 7, at 59 (explaining how this violates professional ethics rules).
  76. .See generally Brown, supra note 41.
  77. .See Gold et al., supra note 12, at 1652–54 (discussing how, unless prosecutorial plea-bargaining leverage is reduced, any attempts to reform the process of plea bargaining can be circumvented).
  78. .For example, many misdemeanor defendants plead guilty not because prosecutors can threaten them with harsher outcomes if they insist on a trial, but rather because of the costs associated with pretrial procedure, such as the costs of multiple pretrial court appearances. See Hessick, supra note 7, at 108–12, 120–27. The pressure to plead guilty in those cases could be reduced by not requiring criminal defendants to repeatedly return to court. Id. at 190–93.
  79. .This claim is addressed in great detail in Hessick, supra note 7.
  80. .See id. at 218 (quoting a public defendant explaining that plea bargaining is “relatively good” for his clients but “really bad for the system”).
  81. .George Fisher, Plea Bargaining’s Triumph: A History of Plea Bargaining in America 31–33 (2003); William Ortman, When Plea Bargaining Became Normal, 100 B.U. L. Rev. 1435, 1457–63 (2020). I am sympathetic to that argument. See Hessick, supra note 7, at 157–64. But it would require a separate review essay to develop the argument in any detail, so I do not address it here.
  82. .Bellin, supra note 2, at 559 (citing Ben Grunwald, Distinguishing Plea Discounts and Trial Penalties, 37 Ga. St. U. L. Rev. 261, 267, 281 (2021)).
  83. .Epstein was credibly accused of sexually exploiting dozens of underage girls; he was given a plea bargain that required him to spend only thirteen months in jail. See Julie K. Brown, Cops Worked to Put a Serial Abuser in Prison. Prosecutors Worked to Cut Him a Break., Miami Herald (Nov. 28, 2018), [].
  84. .See, e.g., Matthew E. Milliken, Cline Defends Fairness of DA’s Office, Herald-Sun, Mar. 14, 2009, 2009 WLNR 4898414 (recounting a prosecutor defending lenient plea bargains for defendants involved in the murder of a child on evidentiary grounds).
  85. .Bellin, supra note 2, at 539.
  86. .See Lafler v. Cooper, 566 U.S. 156, 170 (2012) (noting “the reality that criminal justice today is for the most part a system of pleas, not a system of trials”).