The American criminal justice system is a system of pleas. Few who know it well think it is working. And yet, identifying plausible strategies for law reform proves challenging, given the widely held scholarly assumption that plea bargaining operates “beyond the shadow of the law.” That assumption holds true with respect to substantive and constitutional criminal law—the two most studied bodies of law in the criminal justice system—neither of which significantly regulates prosecutorial power. The assumption is misguided, however, insofar as it fails to account for a third body of law—the subconstitutional law of criminal procedure—that regulates and often establishes the very mechanisms by which prosecutorial plea bargaining power is both generated and deployed.

These hidden regulatory levers are neither theoretical nor abstract. Rather, they exist in strikingly varied forms across our pluralist criminal justice system. This Article excavates these unexamined legal frameworks, conceptualizes their regulatory potential, highlights their heterogeneity across jurisdictions, and exposes the institutional actors most frequently responsible for their content. In so doing, it opens up not only new scholarly terrain but also new potential pathways to criminal justice reform.


    1. Charge Piling: The Law of Joinder and Severance
    2. Case Piling: The Law of Serial Prosecutions
    3. Sentence Piling: The Law of Cumulative Punishment
    1. Factual Overreaching: The Law of Pretrial Evidentiary Review
    2. Legal Overreaching: The Law of Summary Dismissal and Bills of Particulars
    3. Equitable Overreaching: The Law of Equitable Dismissal
    1. Charge Sliding: The Law of Amendment and Dismissal
    2. Verdict Sliding: The Law of Lesser Offenses
    1. The Hydraulics of Reform: Assessing Hidden Law’s Potential Impact
    2. The Politics of Reform: Plea Bargaining’s Hidden Lawmakers



Plea bargaining, we are told, is lawless. It “evolved in the unregulated interstices of our criminal justice system.” 1 Kenneth Kipnis, Criminal Justice and the Negotiated Plea, 86 Ethics 93, 95 (1976); see also William Ortman, Probable Cause Revisited, 68 Stan. L. Rev. 511, 516 (2016) (“American criminal justice backed into plea bargaining, and formal law has long been ambivalent about it.”). And it continues to be driven not by law but by power—the vast, unregulated power of prosecutors. 2 See David Alan Sklansky, The Nature and Function of Prosecutorial Power, 106 J. Crim. L. & Criminology 473, 480–81 (2016) (“The starting point for virtually every discussion of prosecutors in the United States is their . . . ‘almost limitless discretion’ and ‘virtually absolute power.’” (first quoting Erik Luna & Marianne Wade, Prosecutors as Judges, 67 Wash. & Lee L. Rev. 1413, 1414 –15 (2010); then quoting Marc L. Miller, Domination and Dissatisfaction: Prosecutors as Sentencers, 56 Stan. L. Rev. 1211, 1252 (2004))). As plea bargaining scholars have long recounted, prosecutors’ ability to threaten inflated sentences, combined with their power to trade those sentences away for pleas of guilt, allows them to control “who goes to prison and for how long.” 3 William J. Stuntz, Plea Bargaining and Criminal Law’s Disappearing Shadow, 117 Harv. L. Rev. 2548, 2549 (2004) [hereinafter Stuntz, Shadow]; see also infra Part I. As for law, it has abandoned, on this account, its most basic function: channeling prosecutorial power through regulatory constraints. 4 Cf. Kenneth Culp Davis, Discretionary Justice: A Preliminary Inquiry 50–51 (1969) (asserting that “[t]he goal of the rule of law” should be “to distinguish between necessary discretionary power and unnecessary discretionary power,” “to find effective ways to control” the former, and to “cut back” on the latter); Sklansky, supra note 2, at 489 (“[T]he more discretion that prosecutors have, the greater will be the concern, generally speaking, about the power they exercise and vice versa.”). Substantive criminal law, after all, now penalizes so much conduct, so severely, and so many times over that it serves simply to delegate power to prosecutors, transforming them into administrators of an “unwritten criminal ‘law’ that consists only of [their own] discretionary decisions” to charge certain offenses or to offer certain deals. 5 William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 579–80 (2001) [hereinafter Stuntz, Politics]. Meanwhile, the constitutional law of criminal procedure that is ostensibly designed to regulate state power imposes virtually no constraints on prosecutors’ plea bargaining practices at all. 6 See infra Parts II–IV; see also Adriaan Lanni & Carol Steiker, A Thematic Approach to Teaching Criminal Adjudication, 60 St. Louis U. L.J. 463, 469 (2016) (“In a world of guilty pleas, the prosecutor’s determinations of what to charge and what bargain to offer are the ball game, yet the case law regulates this process only minimally.”); cf. Andrew Manuel Crespo, Systemic Facts: Toward Institutional Awareness in Criminal Courts, 129 Harv. L. Rev. 2049, 2054–56 (2016) [hereinafter Crespo, Systemic Facts] (discussing the “regulatory jurisprudence” of constitutional criminal law). Thus, the conventional account: Plea bargaining operates “outside the law’s shadow,” 7 Stuntz, Shadow, supra note 3, at 2558. governed instead only by brute prosecutorial power that is exercised in ways “not usually written down anywhere,” let alone “governed by formal legal standards.” 8 Gerard E. Lynch, Our Administrative System of Criminal Justice, 66 Fordham L. Rev. 2117, 2123 (1998) [hereinafter Lynch, Administrative System].

Against this received wisdom, this Article offers a different account. Plea bargaining, it contends, appears lawless only if by “law” one refers to those two familiar legal pillars of the American criminal justice system—substantive and constitutional criminal law—that together consume aca­demic discussions, in classrooms and in legal scholarship alike. 9 See, e.g., Stephanos Bibas, The Real-World Shift in Criminal Procedure, 93 J. Crim. L. & Criminology 789, 790 (2003) [hereinafter Bibas, Real-World Shift] (book review) (“[A] focus on Supreme Court doctrine continues to rule criminal procedure, in both scholarly articles and casebooks.”); see also infra Parts II–IV (discussing legal scholarship’s consistent focus on constitutional modes of regulating plea bargaining). As for constitutional law’s hold on law school pedagogy, see, for example, Lanni & Steiker, supra note 6, at 464 (describing the “conventional approach” used at “most schools” to teach criminal procedure through a pair of courses focused “on the constitutional constraints on police practices and . . . the constitutional doctrines that structure the adjudicative process”); Ben Trachtenberg, Choosing a Criminal Procedure Casebook: On Lesser Evils and Free Books, 60 St. Louis U. L.J. 543, 544–45 (2016) (explaining that while “the Federal Rules of Civil Procedure guide the syllabus” in civil procedure courses, “the primary law undergirding a criminal procedure syllabus . . . consists of court opinions interpreting the brief constitutional provisions at issue”). Beyond those twin pillars, however, lies a third, unseen but essential body of law that has long been obscured by some of criminal justice scholarship’s most familiar blind spots: It is a creature primarily of state law (not federal law), of court rules and statutes (not constitutional doctrine), and of procedures often seen as relevant only to a bygone era of trial-based litigation (not to the system of pleas that has replaced it). 10 Cf. Sheri Lynn Johnson, Batson Ethics for Prosecutors and Trial Court Judges, 73 Chi.-Kent L. Rev. 475, 477 (1998) (“As has been observed many times, academics tend to focus on appellate courts and cases, perhaps because appellate opinions are so much more accessible than the doings of trial courts, and perhaps because, increasingly, tenure-track academics have little or no personal experience in the trial courts.”); Ronald Wright & Marc Miller, The Screening/Bargaining Tradeoff, 55 Stan. L. Rev. 29, 117 (2002) (criticizing “the legal academy’s . . . ignorance of the wondrous variation in state and local systems”); id. at 55 (asserting that criminal procedure scholars lack “their traditional window on the law” when “issues are not framed in constitutional terms”). And yet, as this Article will show, it is this subconstitutional state law of criminal procedure—the hidden law of plea bargaining—that time and again establishes the mechanisms and legal frameworks through which prosecutorial plea bargaining power is generated and deployed.

Take, for example, charge bargaining, the primary mechanism by which prosecutors control defendants’ sentencing exposure, and with it the so-called trial penalties defendants face if they dare refuse a prosecutor’s invitation to plead guilty. Long criticized as an illicit form of coercion, charge bargaining presents a conundrum under the traditional scholarly account: Given the breadth and depth of substantive criminal law, charge bargaining is routinely diagnosed as a major driver of plea bargaining’s pathology; 11 See infra Part I (unpacking the trial-penalty theory of plea bargaining). For a prominent judicial critique of charge bargaining, see United States v. Kupa, 976 F. Supp. 2d 417, 431–60 (E.D.N.Y. 2013). For prominent scholarly critiques, see, for example, Rachel E. Barkow, Separation of Powers and the Criminal Law, 58 Stan. L. Rev. 989, 1047–50 (2006) [hereinafter Barkow, Separation of Powers]; Stuntz, Politics, supra note 5, at 519–23. But see Frank H. Easterbrook, Plea Bargaining as Compromise, 101 Yale L.J. 1969, 1969–72 (1992). but given prosecutors’ constitutional authority—indeed, their responsibility—to select the charges a defendant will face, it is also seen as an inevitable feature of criminal law’s administration. 12 See infra notes 39, 106 and accompanying text. If one looks beyond substantive and constitutional criminal law, however, charge bargain­ing’s power dynamics become far more complex, for the essential particulars of the practice—ranging from the number of charges the prosecutor can file, to their severity, to their relationship to the defendant’s sentencing exposure, to the ease with which they can be traded away—all directly impact just how much leverage the prosecutor truly has. And crucially, those particulars are in turn dictated by the subconstitutional procedural law of the states—an interlocking set of legal frameworks that comprises the law of joinder and severance, the law of preclusion, the law of cumulative sentencing, the law of pretrial charge review, the law of dismissal and amendment, and the law of lesser offenses. 13 See infra Parts II–IV. The term “subconstitutional” in this Article refers to bodies of law (like those just listed) that lie below federal constitutional law in the hierarchy of legal authorities. Three clarifying points, however, are in order: First, from a regulatory perspective, the frameworks examined in this Article often impose constraints above the so-called constitutional floor and are thus in a sense “supraconstitutional.” Second, the regulatory constraints described here could be imposed by federal constitutional law; they are not, in other words, inherently subconstitutional. But as the following discussion will show, while federal constitutional law could impose such constraints, it consistently has not—and shows little signs of changing course. See infra Parts II–IV. There is thus a need for scholars and reformers to look beyond federal constitutional law to alternative frameworks. See infra section V.B; cf. Andrew Manuel Crespo, Regaining Perspective: Constitutional Criminal Adjudication in the U.S. Supreme Court, 100 Minn. L. Rev. 1985, 1986–88 (2016) [hereinafter Crespo, Regaining Perspective] (discussing “institutional vexations” that may “stymie the Supreme Court’s ability to engage properly and fairly with . . . questions of criminal justice”). Finally, the state procedural law examined here occasionally arises from state constitutional law. As a descriptive matter, however, this occurs only infrequently, which means that plea bargaining’s hidden procedural law is by and large “subconstitutional” in both the federal and state meanings of the term. See infra Appendix (listing sources of state procedural law); cf. infra note 242 (discussing the role of state constitutional law in further detail).

Almost entirely unexamined in existing plea bargaining literature, these hidden legal frameworks address issues that every system of criminal justice administration simply must resolve as it determines, for example, how many charges can be filed in a given case, or whether the sentences associated with those charges will run concurrently or consecutively. And yet, the answers to these inescapable questions, while necessarily shaping prosecutorial power, are neither hardwired nor predetermined. Rather, the choice of which procedural regime to adopt—from a range of potential options—inherently, if often implicitly, presents an important policy choice about how prosecutorial plea bargaining power ought to be structured. 14 The regulatory nature of procedural law is well recognized, and sometimes colorfully captured. See Wisconsin v. Constantineau, 400 U.S. 433, 436 (1971) (“[I]t is procedure that marks much of the difference between rule by law and rule by fiat.”); Regulatory Reform Act: Hearing on H.R. 2327 Before the Subcomm. on Admin. Law & Governmental Relations of the H. Comm. on the Judiciary, 98th Cong. 312 (1983) (statement of Rep. John D. Dingell, Chairman, H. Comm. on Energy & Commerce) (“I’ll let you write the substance . . . and  you let me write the procedure, and I’ll screw you every time.”); see also William J. Stuntz, Privacy’s Problem and the Law of Criminal Procedure, 93 Mich. L. Rev. 1016, 1020 (1995) (“[A]ll procedural rules have substantive effects. . . . [I]t is useful to see how criminal procedure casts its substantive shadow, and how the size and shape of that shadow depends on the interests the law chooses to protect.”).

As Professor Kenneth Culp Davis explained half a century ago, that core question poses one of the central challenges of governmental administration: Too much discretionary power, and “justice may suffer from arbitrariness or inequality”; too little, and “justice may suffer from insufficient individual­izing.” 15 Davis, supra note 4, at 52. In a pluralist criminal justice system such as ours, no fixed formula can prescribe exactly the right amount of prosecutorial power for every jurisdiction, as no two communities confront the same challenges when balancing their citizens’ competing rights to liberty and security. 16 See infra note 210 and accompanying text. And yet, as Davis observed—and as a growing chorus of academics, activists, and politicians from both parties now agree—too often our criminal justice system gets that balance wrong, tolerating an unacceptable excess of prosecutorial power, and with it an unacceptable excess of incarceration, doled out in troublingly unequal ways. 17 See Davis, supra note 4, at 52 (“Americans unquestionably err much more often by making discretion too broad than by making it too narrow . . . .”); see also John F. Pfaff, Locked In 6 (2017) [hereinafter Pfaff, Locked In] (“The primary driver of incarceration is increased prosecutorial toughness when it comes to charging people . . . .”); Sklansky, supra note 2, at 481, 489 (“The concentration of power in the hands of prosecutors has been called the ‘overriding evil’ of American criminal justice . . . .” (quoting Donald A. Dripps, Reinventing Plea Bargaining, in The Future of Criminal Law 55, 60 (Michelle Madden Dempsey et al. eds., 2014))). For a sampling of academic criticism of plea bargaining’s systemic and individual injustices, see Rachel E. Barkow, Institutional Design and the Policing of Prosecutors: Lessons from Administrative Law, 61 Stan. L. Rev. 869, 883–84 (2009) [hereinafter Barkow, Institutional Design] (“The consolidation of adjudicative and enforcement power in a single prosecutor is also troubling because it creates an opportunity for that actor’s prejudices and biases to dictate outcomes. . . . Indeed, researchers have found that, even after controlling for legally relevant factors, race and gender affect charging and sentencing decisions.” (footnote omitted)); Stephanos Bibas, Plea Bargaining Outside the Shadow of Trial, 117 Harv. L. Rev. 2463, 2468 (2004) [hereinafter Bibas, Outside the Shadow] (arguing that “plea bargaining . . . bases sentences in part on wealth, sex, age, education, intelligence, and confidence”); Tracey L. Meares, Rewards for Good Behavior: Influencing Prosecutorial Discretion and Conduct with Financial Incentives, 64 Fordham L. Rev. 851, 878 (1995) (“[T]he prosecutor’s one-sided control of plea bargaining impacts poorer defendants to a greater extent than it impacts wealthier defendants.”); Sonja B. Starr & M. Marit Rehavi, Mandatory Sentencing and Racial Disparity: Assessing the Role of Prosecutors and the Effects of Booker, 123 Yale L.J. 2, 27–31 (2013) (describing racially disparate charging practices); Crystal S. Yang, Free at Last? Judicial Discretion and Racial Disparities in Federal Sentencing, 44 J. Legal Stud. 75, 78 (2015) (same); see also Meares, supra, at 875 (describing “the tendency of some prosecutors to exert the most pressure on defendants in weak cases”); Wright & Miller, supra note 10, at 94 (noting the “urgent” and “substantial problem” of plea bargains ensnaring “the innocent defendant”).

As recognition of these systemic failings turns criminal justice reform into the rallying cry of a generation, the need to understand the regulatory levers hidden within subconstitutional procedural law—and to recognize which institutional actors are responsible for crafting these regulatory regimes—grows only more significant. 18 On the growing momentum in favor of criminal justice reform, see generally Solutions: American Leaders Speak Out on Criminal Justice (Inimai Chettiar & Michael Waldman eds., 2015) (collecting critiques of American overcriminalization from leading politicians, including presidential candidates from both parties); see also Amna A. Akbar, Toward a Radical Imagination of Law, 93 N.Y.U. L. Rev. (forthcoming 2018) (manuscript at 17–30), (on file with the Columbia Law Review) (describing reforms proposed by the Movement for Black Lives). But see Ames C. Grawert & Natasha Camhi, Brennan Ctr. for Justice, Criminal Justice in President Trump’s First 100 Days 1 (2017),
President_Trumps_First_100_Days.pdf [] (describing how the Trump Administration’s criminal justice policy “directly contradicts the emerging consensus among conservatives, progressives, law enforcement, and researchers” and threatens to “set back the national trans-partisan movement to end mass incarceration”). For a thoughtful account of how the so-called “bipartisan consensus” favoring reform masks important ideological divisions, see Benjamin Levin, The Consensus Myth in Criminal Justice Reform, 117 Mich. L. Rev. (forthcoming 2018) (manuscript at 1–5), (on file with the Columbia Law Review).
For in a system of pleas such as ours, in which nearly every sentence of incarceration arises from a plea of guilt, criminal justice reform and plea bargaining reform are of necessity one and the same. 19 See Lafler v. Cooper, 566 U.S. 156, 170 (2012) (“[N]inety-four percent of state convictions are the result of guilty pleas.”). The central goal of this Article is thus to draw this hidden law of plea bargaining into the open, subjecting it for the first time to sustained scholarly analysis. 20 No scholar to date has offered a systematic account of how the various elements of subconstitutional procedural law come together to construct a regulatory framework for plea bargaining power. Indeed, only a small handful of academics recognize the potential significance such law might hold for plea bargaining at all. The most extensive account is Professors Russell Gold, Carissa Hessick, and Andrew Hessick’s recent comparative analysis of how provisions in the Federal Rules of Criminal Procedure and the Federal Rules of Civil Procedure might impact settlement practices in federal court. See Russell M. Gold, Carissa Byrne Hessick & F. Andrew Hessick, Civilizing Criminal Settlements, 97 B.U. L. Rev. 1607, 1639–52 (2017). Professor (now Judge) Stephanos Bibas has also examined how subconstitutional procedural rules might impact plea bargaining, albeit in an essay that (by his own account) trains on a discrete set of “relatively small-bore” reforms. See Stephanos Bibas, Regulating the Plea-Bargaining Market: From Caveat Emptor to Consumer Protection, 99 Calif. L. Rev. 1117, 1153–59 (2011) (urging, inter alia, rules requiring plea agreements to be written down, with key terms in boldface type). Finally, Professor Máximo Langer has offered a theoretical account of plea bargaining coercion that he pairs with a discussion of certain constitutional and subconstitutional doctrines. See Máximo Langer, Rethinking Plea Bargaining: The Practice and Reform of Prosecutorial Adjudication in American Criminal Procedure, 33 Am. J. Crim. L. 223, 270–72 (2006).
In excavating the hidden regulatory levers discussed here, this project joins longstanding scholarly efforts “to identify crucial points in criminal procedure and the criminal justice system that reformers should look at” as they strive to improve how that system operates. Id. at 225, 268; see also Ronald F. Wright, Trial Distortion and the End of Innocence in Federal Criminal Justice, 154 U. Pa. L. Rev. 79, 82 (2005) [hereinafter Wright, Distortion] (urging “a regulatory strategy” to reform plea bargaining, “[b]ecause no one will abolish plea bargains entirely”). See generally Harry T. Edwards, The Growing Disjunction Between Legal Education and the Legal Profession, 91 Mich. L. Rev. 34, 42–43, 55–56 (1992) (promoting “‘practical’ legal scholarship” that analyzes how “complex, existing legal regime[s]” work in order to help “legislators and other policymakers . . . [work toward] law reform”). Existing reform proposals, however, tend to track the scholarship’s narrow focus on substantive and constitutional criminal law—either by proposing reforms internal to those doctrinal arenas, see infra notes 57–58, 71, 85, or by abandoning formal law reform altogether and urging instead various modes of prosecutorial self-regulation. See, e.g., Barkow, Institutional Design, supra note 17, at 896–98; Lynch, Administrative System, supra note 8, at 2143; Meares, supra note 17, at 873; Wright & Miller, supra note 10, at 55–58. For a critique of reform agendas focused solely on law enforcement self-regulation, see Crespo, Systemic Facts, supra note 6, at 2059–65.

That analysis proceeds here in five Parts. Part I begins by deconstruct­ing charge bargaining, the central and most criticized mechanism of prosecutorial power, into its three constituent components, which this Article terms piling on, overreaching, and sliding down. Those components, in turn, form the organizing framework for the three Parts that follow, Parts II, III, and IV, which are the Article’s core and which together offer its two initial contributions: First, they excavate the complex and inter­locking set of procedural levers that serve as charge bargaining’s hidden regulatory framework, analyzing how each lever can be deployed to either facilitate or restrict prosecutorial power, depending on how its accompanying procedural law is structured. Second, by surveying the variety of procedural frameworks employed across the states, the Article exposes the surprising degree of procedural—and thus regulatory—heterogeneity currently in place, and thereby destabilizes assumptions about where the boundaries of feasible reform might lie. 21 This Article intentionally foregrounds subconstitutional state law, to the exclusion of subconstitutional federal law, in an effort to guard against the distorting gravitational pull that federal law routinely exerts on criminal justice scholarship. See Wright & Miller, supra note 10, at 117 (criticizing “the legal academy’s . . . obsessive focus on federal law”); cf. Meares, supra note 17, at 853 n.4 (“[T]he relevant case law of the federal system is simply more accessible for analysis . . . .”). As Professor Bibas writes, a “shift” in focus “from federal to state law broadens criminal procedure scholarship” by encouraging “[s]cholars and students [to] pay more attention to [how] the variety of state procedures . . . may serve different local needs.” Bibas, Real-World Shift, supra note 9, at 803. That said, because subconstitutional procedural law serves a regulatory role in the federal system as well, federal law is reported as a fifty-first exemplar in this Article’s Appendix, infra.

With those expanded horizons of potential reform in mind, Part V reflects on the underlying practical and political forces of the current plea bargaining regime, offering three final contributions: First, it provides a concrete example of what a coordinated suite of subconstitutional procedural reforms might look like. Second, it lays the foundation for future empirical analyses of how these various procedural levers might interact with broader sociolegal forces to impact plea bargaining practices on the ground. Finally, it casts the underlying political economy of plea bargaining in a new light, exposing a previously hidden but ultimately central set of lawmakers: state courts, acting here not in their familiar capacity as adjudicators deciding cases but rather in a fundamentally distinct and ultimately surprising role—as quasi-legislatures, responsible for crafting the heretofore hidden law that governs our criminal justice system of pleas.

I. The Hidden Law of Charge Bargaining

The ability to control a defendant’s sentencing exposure by manipu­lating the charges against him—that is to say, the ability to charge bargain—is widely recognized by scholars as “the core of prosecutorial power in the United States.” 22 Sklansky, supra note 2, at 484–87 (describing scholarly consensus defining the core of prosecutorial power as “the ability to coerce guilty pleas,” which “depends, in turn, on an ability to threaten outcomes . . . [and to] agree to forego charges”). Charge bargaining is not, however, the only mechanism of prosecutorial power, nor the only mechanism structured by subconstitutional procedural law. Prosecutors, for example, can also exercise leverage in plea negotiations by offering to recommend specific sentences in exchange for pleas of guilt, a process known as sentence bargaining. Sentence bargaining is generally seen as less problematic than charge bargaining insofar as it retains a meaningful role for judges. See, e.g., Bibas, Outside the Shadow, supra note 17, at 2534 (“[C]lear sentence bargains are preferable to opaque charge bargains.”); Meares, supra note 17, at 888 (“Limiting charge bargaining necessarily limits the prosecutor’s power because the only tool left to manipulate is a tool over which she must share her power with judges.”); Wright & Miller, supra note 10, at 111–12 (“Sentence bargains do less harm than charge bargains because sentencing decisions necessarily involve many actors.” (emphasis omitted)). For extensions of this Article’s insights to sentence bargaining and other mechanisms of prosecutorial power, see infra note 198. The practice itself is simple enough to describe: A criminal defendant’s sentencing exposure is a function of his likelihood of conviction and his likely sentence if convicted. Those two factors, in turn, are heavily influenced by the charges he faces, which define the possible grounds for conviction, the maximum potential sentence, and frequently the minimum sentence as well. 23 See Bibas, Outside the Shadow, supra note 17, at 2487 (observing that both sentencing guidelines and mandatory minimum “penalties turn on the severity of the offense” of conviction); see also id. (observing that “many state systems have replaced indeterminate sentencing with sentencing guidelines and have adopted mandatory minimum sentences by statute”). Because this Article takes charge bargaining and the attendant manipulation of trial penalties as its point of departure, it focuses on cases in which large sentencing differentials are at issue—namely, felony cases. In some instances, misdemeanor prosecutions raise similar dynamics, particularly when serious collateral consequences such as deportation, loss of government housing, or sex offender registration are at stake. See generally Paul T. Crane, Charging on the Margin, 57 Wm. & Mary L. Rev. 775 (2016) (examining how prosecutors may strategically leverage the collateral consequences associated with misdemeanor offenses). For a discussion of the extralegal forces that may drive outcomes in misdemeanor cases, see infra notes 212–213 and accompanying text. A charge bargain is thus simply an agreement to replace a higher charge with a lower one in exchange for the defendant’s promise to plead guilty, which guarantees the prosecutor a conviction without the expense of trial. 24 Notably, when it comes to avoiding the expenses of litigation, the government routinely bears the vast majority of those costs, given that most defendants are indigent and thus entitled to representation at government expense. See Gideon v. Wainwright, 372 U.S. 335, 343–45 (1963); Thomas H. Cohen, Who Is Better at Defending Criminals? Does Type of Defense Attorney Matter in Terms of Producing Favorable Case Outcomes, 25 Crim. Just. Pol’y Rev. 29, 35 (2014) (reporting defendant indigency rates of 80% in felony cases). But cf. Laura I. Appleman, Nickel and Dimed into Incarceration: Cash-Register Justice in the Criminal System, 57 B.C. L. Rev. 1483, 1485 (2016) (describing “bail fees, translation fees, indigent representation fees, dismissal fees” and other costs paid by indigent defendants).

Yet while such an exchange may sound like an actual bargain, with each party gaining, to quote the Supreme Court, a “mutuality of advantage” from the deal, 25 Bordenkircher v. Hayes, 434 U.S. 357, 363 (1978) (internal quotation marks omitted) (quoting Brady v. United States, 397 U.S. 742, 752 (1970)). most knowledgeable observers describe it as something else: a fundamentally coercive practice (occasionally analogized to torture) that produces involuntary pleas, sometimes to crimes the defendant did not commit. 26 See, e.g., Langer, supra note 20, at 246–47 (describing “coercive” prosecutorial leverage, “involuntary” guilty pleas, and “unilateral adjudication” by prosecutors as “common phenomen[a] in the American criminal justice system”); Lynch, Administrative System, supra note 8, at 2132 (noting that while defense attorneys “may implicitly threaten a trial” there is in truth no actual “exchange of values based on relative bargaining strength,” because prosecutors have “virtually unilateral power to inflict pain on the defendant”); Meares, supra note 17, at 863–66 (rejecting “the Supreme Court’s characterization of plea bargaining as ‘the mutuality of advantage’ by ‘give-and-take’” between parties given that “the prosecutor is able to control the dynamics of plea bargaining” by “controlling the defendant’s exposure to punishment”). For torture analogies, see Donald A. Dripps, Guilt, Innocence, and Due Process of Plea Bargaining, 57 Wm. & Mary L. Rev. 1343, 1343–45 (2016); John H. Langbein, Torture and Plea Bargaining, 46 U. Chi. L. Rev. 3, 3 (1978). On plea bargaining’s innocence problem, see supra note 17 and infra notes 34, 99. The core problem is twofold. First, while defendants always want to minimize their potential sentences, prosecutors rarely want to maximize them, hoping instead to obtain only their preferred sentence, in the most efficient way possible. 27 See Sklansky, supra note 2, at 488 (“The broad consensus among scholars is that prosecutors today are able to bargain for the results they want without giving up much that is important to them, because the outcomes they can credibly threaten under modern sentencing statutes are extraordinarily harsh.”); Stuntz, Shadow, supra note 3, at 2553–54 (“[E]xtra months in prison are not like marginal dollars in civil cases. Once the defendant’s sentence has reached the level the prosecutor prefers . . . adding more time offers no benefit to the prosecutor.”); id. at 2549 (analogizing “criminal law and the law of sentencing” to “items on a menu from which the prosecutor may order as she wishes,” with “no incentive to order the biggest meal possible” but rather simply the incentive “to get whatever meal she wants”); cf. Albert W. Alschuler, The Prosecutor’s Role in Plea Bargaining, 36 U. Chi. L. Rev. 50, 95 (1968) [hereinafter Alschuler, The Prosecutor’s Role] (describing a charge bargain as akin to a prosecutor offering a defendant “the sleeves from his vest”). When a prosecutor does try to maximize incarceration—such as in prosecutions for very serious offenses—her plea bargaining leverage diminishes, and with it the rate of plea bargaining itself. Cf. Lauren M. Ouziel, Legitimacy and Federal Criminal Enforcement Power, 123 Yale L.J. 2236, 2255 (2014) (observing that “[h]omicide offenses have one of the lower guilty plea rates” while also carrying “the highest statutory and Guidelines penalties”). This asymmetry allows prosecutors to trade away “extra” years of incarceration that the defendant desperately wants to avoid but that the prosecutor doesn’t particularly value. As for the second problem: This free leverage is typically overwhelming, because most criminal codes authorize sentences much higher than what a typical prosecutor—or a typical person, for that matter—would actually want to see imposed in a given case. 28 See Sanford H. Kadish et al., Criminal Law and Its Processes 1180 (10th ed. 2017) (“Criminal statutes now commonly permit (or purport to require) draconian punishments that no one expects to be imposed in the typical case.”); Barkow, Institutional Design, supra note 17, at 880–81 (“[Legislators] . . . routinely pass[] laws with punishments greater than the facts of the offense would demand to allow prosecutors to use the excessive punishments as bargaining chips . . . to obtain . . . the more appropriate sentence via a plea instead of a trial.”); Adriaan Lanni, The Future of Community Justice, 40 Harv. C.R.-C.L. L. Rev. 359, 388 (2005) (“[W]hen given detailed descriptions of specific cases, studies show that [survey] respondents often suggest sentences that are more lenient than the mandatory minimum in their jurisdiction.”); see also Barkow, Separation of Powers, supra note 11, at 1034 (“[T]hose who do take their case to trial and lose [in such a system] receive longer sentences than even Congress or the prosecutor might think appropriate, because the longer sentences exist on the books largely for bargaining purposes.”). For a striking example of extreme sentencing leverage, consider Bordenkircher v. Hayes, a case in which a defendant accused of writing a bad check for $88.30 received a mandatory life sentence after rejecting a plea deal carrying a sentence of five years. 434 U.S. at 358–59. Thus, by threatening a seriously inflated set of charges and then offering to replace it with the charges that she truly desires, the prosecutor is able to control the defendant’s incentive to plead guilty, and with it the outcome of any subsequent “negotiation.” 29 See, e.g., Meares, supra note 17, at 861 (“[P]rosecutorial discretion . . . allows the prosecutor to manipulate the charging decision to control the defendant’s decision to plead guilty or go to trial.”). In the aggregate, prosecutors so empowered can obtain more convictions, with longer sentences, at lower costs—all preconditions for mass incarceration. 30 See Albert W. Alschuler, Lafler and Frye : Two Small Band-Aids for a Festering Wound, 51 Duq. L. Rev. 673, 705 (2013) (“By lowering the price of imposing criminal punishment, plea bargaining gave America more of it.”); John F. Pfaff, The Micro and Macro Causes of Prison Growth, 28 Ga. St. U. L. Rev. 1239, 1254 (2012) (“Prison growth has been driven by admissions, and at least since the early 1990s admissions have been driven by prosecutorial filing decisions.”); see also William J. Stuntz, The Collapse of American Criminal Justice 258 (2011) (observing that unchecked plea bargaining leverage “allow[s] the government to do two things that, in combination, [are otherwise] hard to pull off: raise the guilty plea rate and raise average sentences, at the same time”). For other factors contributing to mass incarceration, see infra notes 212–213 and accompanying text.

In practice, charge manipulation involves three interrelated moves. First, the prosecutor can inflate the quantity of charges the defendant faces, by piling on overlapping, largely duplicative offenses—increasing with each new charge the defendant’s potential sentence, his risk of conviction, and the “sticker shock” of intimidation that accompanies a hefty charging instrument. 31 “Piling on” is also sometimes known as “charge-stacking,” Stuntz, Politics, supra note 5, at 594, “horizontal overcharging,” Meares, supra note 17, at 868 (citing Alschuler, The Prosecutor’s Role, supra note 27, at 85–88), or “redundant charging,” Michael L. Seigel & Christopher Slobogin, Prosecuting Martha: Federal Prosecutorial Power and the Need for a Law of Counts, 109 Penn St. L. Rev. 1107, 1121 (2005). On the connection between charge-stacking and increased sentence exposure, see infra section II.C (discussing cumulative sentencing). On the connection between charge-stacking and increased risk of conviction, see Edith Greene & Elizabeth F. Loftus, When Crimes Are Joined at Trial, 9 Law & Hum. Behav. 193, 194 (1985) (discussing an experiment capturing a “‘halo’ effect” that causes jurors to view a defendant more negatively when he is charged with multiple offenses); Andrew D. Leipold, How the Pretrial Process Contributes to Wrongful Convictions, 42 Am. Crim. L. Rev. 1123, 1143 (2005) (observing that in some instances “[a] defendant who is guilty of one charge but innocent of another may find it difficult” to defend successfully against both at once); Andrew D. Leipold & Hossein A. Abbasi, The Impact of Joinder and Severance on Federal Criminal Cases: An Empirical Study, 59 Vand. L. Rev. 349, 355 (2006) (“[T]he more counts in the indictment, the quicker the jury may be to assume that the accused must be guilty of something.”); Seigel & Slobogin, supra, at 1125–26 (noting that multiple charges increase the odds of a “horse-trade” or a “compromise verdict of guilty”). On the connection between charge-stacking and defendant intimidation, see Bibas, Outside the Shadow, supra note 17, at 2518–19 (discussing the psychological anchoring effect that renders defendants initially charged with very high offenses “more likely to think that they are getting good deals when they are offered lower sentences”); Seigel & Slobogin, supra, at 1126 (“[M]ultiple charges intimidate defendants.”). Finally, note that charge-stacking may also prompt less-than-scrupulous defense attorneys to “tell their clients that their superior negotiating skills forced the prosecutor to ‘drop some charges,’” even if the dropped charges would not actually “amount to a shorter sentence.” Seigel & Slobogin, supra, at 1127 (“Prosecutors are surely aware of this dynamic, and may use redundant charging to take advantage.”); cf. Eve Brensike Primus, Culture as a Structural Problem in Indigent Defense, 100 Minn. L. Rev. 1769, 1769 (2016) (observing that “too many lawyers appointed to represent poor criminal defendants do not perform their intended role in the system, because they have been conditioned not to fight for their clients”). Second, the prosecutor can achieve similar effects by inflating the substance of the charges themselves, overreaching beyond what the law, the evidence, or the equities of the case support. 32 Cf. Meares, supra note 17, at 865 n.47, 868 (citing Alschuler, The Prosecutor’s Role, supra note 27, at 85–88) (using the term “vertical overcharging”). Finally, after deploying these tactics to “jack up the threat value of trial,” 33 Stuntz, Politics, supra note 5, at 594. the prosecutor can capitalize on the ensuing leverage by sliding down from her initial threat to the lower set of charges that she actually prefers. Indeed, it is the difference between the threat and the subsequent offer that constitutes the prosecutor’s power: The larger the differential, the more likely the defendant is to plead guilty—whether he is in fact guilty or not. 34 See Langer, supra note 20, at 229 (summarizing scholarship finding prosecutors can create a “sentence differential [that] leaves defendants with no rational choice but to plead guilty,” simply “by offering a sentence substantially lower than the one expected at trial”); Wright, Distortion, supra note 20, at 109 (“The difference between the predicted sentence after a trial conviction and the predicted sentence after a guilty plea could become so large that some defendants would not accurately weigh their options and would not dare go to trial, even with a strong defense.”); Hans Zeisel, The Offer that Cannot Be Refused, in The Criminal Justice System 558, 559–60 (Franklin E. Zimring & Richard S. Frase eds., 1980) (“[T]he greater the difference between the offered sentence and the sentence expected after conviction at trial, the more defendants will plead guilty and avoid trial.”); see also Oren Gazal-Ayal, Partial Ban on Plea Bargains, 27 Cardozo L. Rev. 2295, 2304 (2006) (“[P]rosecutors can extract a guilty plea in almost any case, regardless of the real culpability of the defendant . . . [because even] innocent defendants are willing to accept minor punishment in return for avoiding the risk of a much harsher trial result.”); id. (“Only very rarely is the highest acceptable sentence [to] a defendant zero . . . .”). Of course, as the Supreme Court has recently observed, the dynamic works in the opposite direction as well: The smaller the differential, the more likely the defendant will be to insist on his right to trial. See Lee v. United States, 137 S. Ct. 1958, 1966–67 (2017) (“[When] the respective consequences of a conviction after trial and by plea . . . are . . . similarly dire, even the smallest chance of success at trial may look attractive. For example, a defendant with no realistic defense to a charge carrying a 20-year sentence may nevertheless choose trial, if the prosecution’s plea offer is 18 years.”).

To make these three moves more concrete, consider a straightforward example, to which we will return throughout the discussion to follow: Imagine a defendant suspected of approaching someone on a street corner at night, of pointing a gun at that person, of ordering them to move a few steps to the left (out from under a streetlamp), and, finally, of taking their wallet and running off with it. To any lay observer, the crime alleged here is straightforward: armed robbery. And yet, in practice, a prosecutor could and routinely would commence a prosecution against such a defendant by piling on a host of additional charges, including (to list just some examples) aggravated assault, theft, threats, possession of a weapon, and using a firearm during a crime of violence. 35 Cf. infra Part II (describing charge piling). Moreover, given the defendant’s alleged command to move out from under the streetlamp, the prosecutor might also overreach, tacking on the far more serious but questionably applicable charge of kidnapping for good measure. 36 Cf. infra section III.B and note 148 (discussing legal overreach and the law of kidnapping). Finally, bringing her leverage to bear, the prosecutor would then offer to slide down from these inflated charges to the charge that she—and she alone—deems appropriate, based on her personal assessment of the evidence and of the defendant’s culpability: Plead guilty to a single count of armed robbery, she tells the defendant—or, even more enticingly, to misdemeanor counts of theft and possession of a weapon—and everything else will go away. 37 To be sure, the defendant who accepts the deal is better off than he would be had he gone to trial and lost. But so long as the sentence associated with the inflated charges is one that no one actually wants him to serve, see supra notes 26–29, then his supposed “deal” isn’t really much of a bargain. See Gerard E. Lynch, Screening Versus Plea Bargaining: Exactly What Are We Trading Off?, 55 Stan. L. Rev. 1399, 1402–04 (2003) (doubting whether defendants are “the beneficiaries of ‘bargains’” in the conventional sense of the word). And of course, in pleading guilty, the defendant loses any chance of being acquitted.

As plea bargaining scholars consistently recognize, prosecutorial charging discretion exercised in this fashion “translates into power in the plea bargaining context.” 38 Meares, supra note 17, at 863. And yet this is also the point at which plea bargaining scholarship starts to run out, for it frequently assumes that such power is an intractable feature of criminal justice administration in a system that does not (and arguably cannot) tell prosecutors what charges to file in a given case. 39 Id. at 863–66 (asserting that “prosecutorial discretion at the charging stage undeniably . . . allows the prosecutor to control, essentially unilaterally, the defendant’s ability to plead guilty in most cases,” and describing such power as both “vast” and “well-entrenched”); see also infra note 170 and accompanying text. But cf. Rebecca Roiphe, The Duty to Charge in Police Use of Excessive Force Cases, 65 Clev. St. L. Rev. 503, 508–09 (2017) (noting that a few “states provide some mechanism for [private] citizens to initiate charges with approval of a court”). The true scope of prosecutors’ charge-bargaining power, however, is contingent on much more than just the decision of what charges to file, turning instead on a broader set of procedural questions: With respect to piling on, how many charges can a prosecutor threaten and how much will each additional charge increase the defendant’s sentencing exposure? With respect to overreaching, what standards must be satisfied before substantively inflated charges can proceed beyond the filing stage, and how will those standards be enforced? With respect to sliding down, what restrictions, if any, will be placed on the prosecutor’s ability to replace one set of charges with another?

Too often, these essential questions go overlooked—perhaps because their answers are found not in the familiar constitutional law of criminal procedure but rather in its unexamined counterpart: the subconstitutional procedural law of the states. The following three Parts excavate and analyze that unexamined body of law, mapping for each of the three charge-bargaining components just described the mechanisms by which prosecutors exercise their authority and the procedural law through which that authority is structured. The following table gives an overview of the discussion to come.

Table 1: Overview of Charge-Bargaining Components, Implementing Mechanisms, and Regulatory Frameworks

Part Charge-Bargaining Component Implementing Mechanism Regulatory Framework
(“The Law of . . .”)
II. Piling On Charge Piling Joinder and Severance
Case Piling Preclusion
Sentence Piling Cumulative Sentencing
III. Overreaching Factual Overreaching Pretrial Evidentiary Review
Legal Overreaching Summary Dismissal and
Bills of Particulars
Equitable Overreaching Equitable Dismissal
IV. Sliding Down Charge Sliding Amendment and Dismissal
Verdict Sliding Lesser Offenses

II. Piling On

While prosecutors likely pile charges on top of each other in the majority of cases that they file, their ability to do so is not some inherent feature of their authority. 40 Cf. Leipold & Abbasi, supra note 31, at 351 (“[M]ore than half of all federal defendants are charged with multiple counts . . . .”). Rather, the ability to pile on charges is simply the ability to join multiple charges together in a single case—an ability that the law of joinder, together with its fellow traveler, the law of severance, both establishes and defines. And while joinder and severance determine how many charges can be stacked on top of each other, a prosecutor’s leverage is also affected by the number of piles (that is, the number of separate prosecutions) she can threaten, and by each pile’s potential “weight,” as measured in years in prison—two issues governed by the law of preclusion and the law of cumulative sentencing. This Part examines these interlocking procedural frameworks in turn.

A. Charge Piling: The Law of Joinder and Severance

The law of joinder answers the most basic question that any analysis of prosecutorial charge piling must resolve: How many charges can a prosecutor pile on in a single case? To our hypothetical armed robbery defendant, the question is significant: Will he face potentially dozens of criminal charges arising from that one alleged crime? 41 See supra text accompanying note 35 (offering a list of potential charges); see also Andy Grimm, Shomari Legghette Formally Indicted for Cmdr. Paul Bauer’s Murder, Chi. Sun Times (Mar. 9, 2018),
formally-indicted-for-cmdr-paul-bauers-murder [] (describing a case in which a defendant accused of a single homicidal act was charged with “more than two dozen counts of murder, as well as a combined 32 additional counts of armed violence and related weapons charges—a total of 56 counts in all”); id. (describing this charging practice as “not uncommon”).
Might other alleged robberies, including some in which the evidence against him is quite weak, be added to the pile? 42 Cf. Leipold, supra note 31, at 1143 (describing the difficulty of defending against multiple accusations that implicate different underlying evidence and potentially different defense strategies). Might entirely unrelated accusations (say, of drug possession) be thrown onto the pile as well? 43 See infra notes 47–48 and accompanying text (discussing similar-offense and unlimited joinder).

Rarely examined in courses or scholarship concerning criminal procedure, the law of joinder answers each of these questions—in different ways across jurisdictions. 44 Many criminal procedure courses do not discuss joinder at all, see, e.g., Lanni & Steiker, supra note 6, at 473–79 (providing a model syllabus), and casebooks give the topic short shrift as well, discussing it only briefly and with near-exclusive focus on the potential trial consequences of federal joinder rules, see, e.g., Ronald Jay Allen et al., Comprehensive Criminal Procedure 1168–83 (4th ed. 2016); Joshua Dressler & George C. Thomas III, Criminal Procedure: Principles, Policies and Perspectives 912–19 (4th ed. 2010); Yale Kamisar et al., Advanced Criminal Procedure: Cases, Comments and Questions 1062–89 (14th ed. 2015). As for scholarship, the treatment described as “[t]he best analysis of joinder and severance” is now forty years old and focuses on the trial consequences of joining together multiple defendants, not the plea bargaining consequences of joining multiple charges. See Leipold & Abbasi, supra note 31, at 351 n.2 (discussing Robert O. Dawson, Joint Trials of Defendants in Criminal Cases: An Analysis of Efficiencies and Prejudices, 77 Mich. L. Rev. 1379 (1979)). Other works in the field include a trio of short student notes of similarly dated vintage, see Samuel A. Baron, Note, A Look at the Tennessee Multiple Offender and the Joinder and Severance of Criminal Offenses for Trial, 7 Memphis St. U. L. Rev. 457 (1977); James Farrin, Note, Rethinking Criminal Joinder: An Analysis of the Empirical Research and Its Implications for Justice, 52 Law & Contemp. Prob. 325 (1989); Note, Joint and Single Trials Under Rules 8 and 14 of the Federal Rules of Criminal Procedure, 74 Yale L.J. 553 (1965), and a handful of articles assessing joinder in terms of its potential trial consequences, see Kenneth S. Bordens & Irwin A. Horowitz, Joinder of Criminal Offenses: A Review of the Legal and Psychological Literature, 9 Law & Hum. Behav. 338, 343–49 (1985); Leipold, supra note 31, at 1142–47; Leipold & Abbasi, supra note 31, at 355–56. Meanwhile, some now-dated treatments from beyond legal scholarship misdescribe joinder as an issue affecting only a “limited number of cases” and involving only a “rudimentary” legal framework. Standards for Criminal Justice: Joinder & Severance, ch. 13, intro. note at 13.4 (Am. Bar Ass’n 1980) (asserting that the law of joinder “often consist[s] of no more than a consignment of [the issue] to the unguided and largely unreviewable discretion of the trial judge”). And in so doing it offers a diverse range of regulatory models, each of which can be arrayed along a spectrum, from those that maximize prosecutorial charge-piling power to those that restrict or even eliminate it. 45 The conceptual schematics offered throughout this Article are largely inductive heuristic devices, derived from the actual procedural regimes employed across the states. As such, they present and conceptualize a diverse array of regulatory models, without purporting to capture every possible model imaginable.

Figure 1: Conceptual Frameworks for the Law of Joinder


In this schematic, regulatory constraints on prosecutorial power increase as the models proceed from left to right, such that procedural regimes maximizing prosecutorial power appear on the left while regimes minimizing such power appear on the right. Thus, at one pole of the spectrum (the leftmost above), an unlimited-joinder regime affords prosecutors maximal authority to charge in one case any and all offenses allegedly committed by the same defendant. Conversely, at the opposite pole, a nonjoinder regime would permit only a single charge per case, thereby eliminating charge piling altogether and ultimately forcing prosecutors to make a significant choice: Either forgo additional charges or pursue them in a resource-intensive string of separate prosecutions. 46 See infra section II.B (discussing serial prosecutions in detail). Finally, in between these poles, a series of fact-based joinder regimes can be arrayed along a subsidiary spectrum, with each model tying prosecutors’ joinder authority to the relationship between the charges themselves. Thus, joinder might be permitted: (a) most broadly, for all offenses that pertain to a similar type of crime (for example, all crimes of violence or all property crimes allegedly committed by the same defendant, even if committed miles and years apart and sharing no factual connection); 47 Cf. Baron, supra note 44, at 458 (describing similar-offense joinder as sometimes permitting joinder of all “crimes involving the taking of property [or all] crimes of violence against another person”); Joinder & Severance, 44 Geo. L.J. Ann. Rev. Crim. Proc. 361–62 n.984 (2015) (citing cases defining the scope of a “similar” offense). (b) more narrowly, for offenses that are part of a common plan or scheme (for example, a conspiracy to commit a series of robberies); (c) more narrowly still, for offenses that are part of the same chain of events (for example, a string of robberies allegedly committed over the span of an hour); or (d) most narrowly, for offenses that arise from a single factual incident (for example, a single robbery). 48 Cf. Standards for Criminal Justice: Joinder and Severance, supra note 44, § 13-1.2 & cmt., § 13-1.3 & cmt., § 13-2.1 & cmt. (describing various joinder models).

Regardless which of these regimes one ultimately adopts, the choice itself clearly impacts just how high a prosecutor’s pile of charges can grow—or whether there can be a pile at all. Prosecutors’ charge-stacking authority, however, can also be affected by the closely related law of severance, which can mitigate prosecutors’ joinder authority by granting judges or defendants some power to divide up the prosecutor’s chosen pile. 49 Note that the power to sever charges differs from the power to enforce the extant joinder rules. In the latter circumstance, the defendant can assert “misjoinder” to break up the charges. See generally Joinder & Severance, supra note 47, at 372–78. Severance, by contrast, allows the defendant or the judge to divide charges that are joinable under the applicable joinder regime. And here, too, different procedural regimes can fall along a spectrum based on how much they facilitate or restrict prosecutorial power.

Figure 2: Conceptual Frameworks for the Law of Severance


At one pole of this spectrum (again, the leftmost above) there is simply no severance mechanism at all, and thus no restraint on charge stacking beyond that afforded by the underlying joinder regime. Alternatively, moving one spectrum position to the right, judges might be granted some reservoir of equitable authority to shrink a prosecutor’s initial pile of charges, either according to their discretion or pursuant to some harm-balancing analysis—an approach that will constrain prosecutorial authority to whatever extent judges use their authority to do so. 50 See infra section V.B (discussing the relationship between prosecutorial constraint and judicial discretion); cf. Leipold & Abbasi, supra note 31, at 361 (suggesting that courts are “miserly in granting severance”); id. at 361 n.55 (“[D]efendants generally have not fared very well under rules and statutes which permit them to obtain a severance of offenses only upon proof of prejudice.” (alteration in original) (internal quotation marks omitted) (quoting 4 Wayne R. LaFave et al., Criminal Procedure § 17.1(f), at 602 (2d ed. 1999))). Equitable-failsafe regimes come in different stripes. As the discussion in the text suggests, one approach is to grant broad discretion to the judge. See, e.g., Zafiro v. United States, 506 U.S. 534, 541 (1993) (holding that the federal severance rule “leaves the determination . . . to the sound discretion of the district courts”). Another is to require some sort of balancing inquiry that weighs the prejudice of joinder and of severance to the respective parties, see, e.g., Mo. Sup. Ct. R. 24.07 (authorizing severance if “[a] party makes a particularized showing of substantial prejudice”), or that weighs potential prejudice to the defendant against competing values, such as judicial economy, see, e.g., State v. Bythrow, 790 P.2d 154, 156 (Wash. 1990). It is also possible to fold rule-like corollaries into an equitable-failsafe regime, deeming severance presumptively warranted, for example, if “evidence of one crime [would be] inadmissible” at the trial for the other. Drew v. United States, 331 F.2d 85, 89 (D.C. Cir. 1964) (applying a rebuttable presumption in favor of severance in such circumstances). Finally, severance-based restrictions can offer their greatest check on prosecutorial power by affording defendants themselves a right to sever charges, either in the form of a partial veto that entitles them to sever charges that are too attenuated, or in the form of an absolute right to divide the prosecutor’s pile as they see fit.

Given the close functional relationship between joinder and severance, it is ultimately the interaction between these two bodies of law that defines the full scope of prosecutorial charge-piling power: The breadth of prosecu­torial joinder power, as constrained by judges’ or defendants’ severance power, is the key variable of interest. Thus, a third and final conceptual schematic blends the two doctrinal frameworks together.

Figure 3: Conceptual Frameworks for Joinder and Severance Combined


Here, the spectrum’s leftmost pole is home to an unlimited joinder rule that is constrained, if at all, by only an equitable-severance failsafe, without any corresponding severance as of right for defendants. The spectrum then progresses through a range of fact-based joinder–severance permutations, each of which could be composed of a joinder rule paired (or not) with a partially offsetting severance-as-of-right rule. 51 These models could also be supplemented (or not) by a form of equitable severance. To see how different permutations of joinder and severance-as-of-right rules could give rise to functionally equivalent joinder–severance regimes, consider a prosecutor who can join A and B, and can also join C subject to the defendant’s veto. Such a prosecutor has the same effective power as a prosecutor who can join only A and B but who is not subject to a severance veto, or a prosecutor who can join A, B, C, and D subject to a veto of C and D. Finally, at the spectrum’s rightmost pole, prosecutorial power reaches its nadir in the form of either a nonjoinder regime or a universal-severance-as-of-right regime, either of which produces a similar functional result: a world in which no defendant can ever be forced to confront one charge piled on top of another. 52 The key conceptual difference between a nonjoinder regime, in which prosecutors simply cannot file more than one charge, and a universal-severance-as-of-right regime, in which the defendant can separate any charges joined together, is that the defendant’s veto power under the latter model could itself potentially be traded away in the bargaining process. See infra note 211 (discussing “procedure bargaining”). Additionally, a defendant’s request for severance might be deemed a waiver of any subsequent attempt to block serial prosecutions. See infra section II.B; cf. Currier v. Virginia, 138 S. Ct. 355, 355 (2017) (granting certiorari to consider whether a defendant can invoke the issue-preclusive protection of the Double Jeopardy Clause if he previously consented to the severance of multiple charges into separate prosecutions).

Though presented thus far as conceptual abstractions, the procedural regimes suggested above—and throughout the Article to follow—are in fact derived from procedural frameworks currently employed across the fifty states. That state-by-state diversity is captured in the chart below, which reports the results of a comprehensive survey that identifies each state’s joinder regime and severance regime and assesses their interaction to produce each state’s composite joinder–severance classification.

Figure 4: Variability Across States in the Law of Joinder–Severance 53 For sources, see infra Appendix at Table A.


As this summary makes clear, the range of existing joinder–severance regimes essentially spans the waterfront, from the maximal to the minimal poles of prosecutorial power. To be sure, most states fall toward the broader end of the spectrum, with twenty-six permitting joinder even for factually unrelated offenses, so long as those offenses are similar in kind. 54 Cf. supra note 47 (discussing the definitional scope of “similar” in this context). And yet, reformers hoping for greater restraints on prosecutorial power will be encouraged to discover a broad severance-as-of-right regime alive and well in the wild, so to speak, over at the spectrum’s far right end. 55 See Tex. Penal Code § 3.04(a) (2017) (“Whenever two or more offenses have been consolidated or joined . . . the defendant shall have a right to a severance of the offenses.”). Notably, while Texas initially embraced a truly universal-severance-as-of-right regime when it first adopted its current framework, see 1973 Tex. Gen. Laws 883, 891, it has since added a series of enumerated exceptions that deny the right for specific crimes related to drunk driving, gang-related offenses, sexual assault, and crimes against vulnerable persons, see 1997 Tex. Gen. Laws 2252; 2005 Tex. Gen. Laws 1429 – 30 (codified at Tex. Penal Code §§ 3.03(b), 3.04(a)); see also Tex. Penal Code § 3.04(b) (providing that if the defendant exercises severance as of right he forfeits the potentially otherwise applicable guarantee of concurrent sentencing set forth in Tex. Penal Code § 3.03); cf. infra section II.C (discussing concurrent sentencing).
Looking beyond Texas, a nonjoinder regime was in place in Arkansas up until 2005. See Ark. Code Ann. § 16 - 85 - 404 (1987) (“An indictment . . . must charge but one . . . offense . . . .”), repealed by Arkansas Criminal Code Revision Commission’s Bill, Act 1994, § 503, 2005 Ark. Acts 6932, 7459–61; see also Ark. Code Ann. § 16 - 85 - 404 (1987) (enumerating offense-based exceptions). Oklahoma has flirted with nonjoinder as well. Compare Okla. Stat. Ann. tit. 22, § 404 (West 2016), and Kramer v. State, 257 P.2d 521, 523 (Okla. Crim. App. 1953) (stating that consolidating multiple charges for a single trial “is not approved” and should be permitted “only in exceptional circumstances where the defendant requests it”), with Sanders v. State, 612 P.2d 1363, 1366 (Okla. Crim. App. 1980) (“[J]oinder . . . lies within the discretion of the trial court.”), and Dodson v. State, 562 P.2d 916, 923 (Okla. Crim. App. 1977) (Brett, J., concurring specially) (“[T]here is no doubt . . . that it was the intention of the legislature to authorize . . . the joinder for trial of two or more indictments or informations, or both.”).

Indeed, that particular specimen (Texas) takes an approach to joinder and severance not unlike a reform proposal advanced by Professors Michael Seigel and Christopher Slobogin, who have urged courts to constrain “the power of the prosecution to charge multiple, overlapping, and redundant crimes” by barring prosecutors from filing more than one charge for any given criminal event. 56 Seigel & Slobogin, supra note 31, at 1128. Texas restricts prosecutorial charge piling through a severance-as-of-right framework as opposed to the essentially nonjoinder-based “law of counts” that Seigel and Slobogin propose, id., and its regime contains some exceptions that make it somewhat less robust than theirs. See supra note 55 (describing exceptions to the Texas regime). And yet, while Seigel and Slobogin understand well the relationship between joinder and prosecutorial power, their essay also exemplifies a lacuna that often afflicts plea bargaining scholarship: They do not examine the procedural frameworks actually employed on the ground, and as a result propose a reform agenda grounded primarily in constitutional law—in this case, a proposal to peg a nonjoinder rule to either “the due process clause” or “separation of powers.” 57 Seigel & Slobogin, supra note 31, at 1128–29. Seigel and Slobogin also briefly acknowledge the possibility that courts might prompt such reform through their general “common law power,” id., though they do not engage the mechanism through which courts primarily construct subconstitutional law: rulemaking, see infra Part V.B. Aside from courts, Seigel and Slobogin also briefly—and skeptically—raise the possibility that Congress might reform the law of joinder. Seigel & Slobogin, supra note 31, at 1128–29. However, like many criminal procedure scholars, they focus on the federal system, and thus do not engage the robust and diverse law of joinder operating at the state level.

Seigel and Slobogin are not alone in looking to constitutional remedies for constraints on prosecutors’ charge-piling power. 58 See, e.g., Stuntz, Politics, supra note 5, at 594 (asserting that courts should address prosecutorial overcharging by “reconfigur[ing] double jeopardy law . . . to limit [prosecutors’] power to pile on separate offenses”). But law reform dependent on constitutional law in this arena, by its proponents’ own admission, faces a steep uphill battle, given the Supreme Court’s steadfast refusal to entertain any such doctrinal revisions. 59 See id. at 594–96 (acknowledging that “the legal case is not . . . strong” (emphasis omitted)). The closest that constitutional law comes to addressing charge stacking is the modest restriction it imposes against multiple punishments, a prohibition enforced only on the back end at sentencing, not at the front end where a potentially unlimited stack of charges can maximize prosecutorial leverage ex ante. Even that back-end protection, moreover, is quite limited, as explained infra section II.C. Law reform aimed at charge stacking, however, need not—and ought not—focus exclusively on that one constitutional lever, because joinder and severance are not primarily creatures of constitutional law. Rather, as the analysis here shows, the size of the pile of charges that a prosecutor can threaten in any given case is both established and defined by the subconstitutional law of joinder and severance, a heretofore hidden law of plea bargaining constructed in the first instance—and in various different ways—by each of the fifty states.

B. Case Piling: The Law of Serial Prosecutions

If a state deploys its law of joinder and severance to restrict prosecutors’ charge-piling power, what might follow? One possibility is that prosecutors will employ a hydraulic counter-response, threatening defendants with a string of separate prosecutions that charges separately what the joinder regime prohibits charging all together. Recall our hypothetical armed robbery defendant. In a permissive-joinder regime, he might be charged not only with armed robbery but also with aggravated assault, theft, threats, unlawful possession of a weapon, brandishing a firearm, and so on and so forth, until one criminal episode yields potentially dozens of charges. If, however, the extant joinder regime permits only one charge per case, the prosecutor might instead threaten to file a seemingly endless string of separate prosecutions—an approach that may actually leave the defendant worse off, as his aggregate risk of conviction rises with each new case added to the string.

Defendants confronting such a case-piling tactic have two principal protections, one practical, the other grounded in its own body of procedural law. The practical protection stems from the fact that prosecutions are costly affairs—and that defendants have only one body to jail. Given these realities, a string of prosecutions will unfold in one of two ways. Either the defendant will be convicted in the first case, after which the prosecutor will have little incentive to continue pursuing the others; or he won’t be convicted in the first case, at which point the prosecutor will have to decide whether to invest precious resources in a second trial, after having just received a public signal that at least some of her charges are weak. And because defendants, aided by counsel, can perceive this dynamic ex ante, the prosecutor’s initial threat to file multiple cases may ring hollow, as the defendant will know that the prosecutor could very well lack the resources to follow through. 60 Cf. United States v. Dixon, 509 U.S. 688, 710 –11 n.15 (1993) (observing that the “press of other demands upon prosecutorial and judicial resources” will reduce prosecutors’ incentives to pursue “repeated prosecutions of a single offender”); Elizabeth T. Lear, Contemplating the Successive Prosecution Phenomenon in the Federal System, 85 J. Crim. L. & Criminology 625, 648 (1995) (“[Prosecutors] juggle impressive caseloads. Every duplicative prosecution means that they must ignore, abandon, or downgrade another case.”); Leipold & Abbasi, supra note 31, at 393 (“There will be times when the prosecutor has enough evidence to bring an additional charge [against the defendant] . . . but because of resource constraints will only do so if the new charge . . . can be joined with others.”); Seigel & Slobogin, supra note 31, at 1122–23 (noting “[m]ost prosecutors would not want” to pursue “several different trials” against a defendant “simply because of efficiency concerns”).

But protections grounded in resource constraints can be fickle, as prosecutors’ resources and their decisions about how to allocate them can vary from office to office, or from case to case. 61 Of course, more often than not, prosecutorial resource constraints are substantial, which is why joinder and severance regimes are imbued with regulatory effect in the first place: Together these procedural levers can reduce the cost of prosecuting multiple charges by permitting the prosecutor to bundle them together, or they can increase those costs by requiring separate, resource-intensive prosecutions for each. As Professors Richard Bierschbach and Stephanos Bibas observe, such resource constraints can be “a feature, not a bug” of criminal justice administration, insofar as they impose restraints on prosecution and in turn restraints on (over) incarceration. Richard A. Bierschbach & Stephanos Bibas, Rationing Criminal Justice, 116 Mich. L. Rev. 187, 233 (2017); see also Stephanos Bibas, Sacrificing Quantity for Quality: Better Focusing Prosecutors’ Scarce Resources, 106 Nw. U. L. Rev. Colloquy 138, 139 (2011),
cgi/viewcontent.cgi?article=1061&context=nulr_online (on file with the Columbia Law Review) (“In a world of overcriminalization, limited budgets are not all bad. The silver lining is that . . . [r]esource constraints and scarcity can force prosecutors to rank priorities, mitigating in practice the problem of overcriminalization on the books.”); cf. infra text accompanying notes 209–210 (discussing inevitable criminal justice tradeoffs).
The ability to threaten a series of separate prosecutions, however, is also constrained by law—specifi­cally, by the law of preclusion, which determines whether a prosecutor’s decision to file one case will block her from filing some subsequent set of cases against the same defendant. Typically analyzed through only the narrow constitutional lens of the Double Jeopardy Clause, 62 See infra note 71. the law of preclusion in fact entails a broader, largely unexamined, subconstitutional framework, which addresses two related regulatory questions: First, when will one case be deemed to preclude another? And second, will the law of preclusion, combined with the law of joinder, force prosecutors to winnow down the charges otherwise available to them under the substantive criminal law of the jurisdiction?

As to the first of these issues, note that the initial prosecution’s preclusive effect is inversely related to the prosecutor’s power to bring later charges: The broader the preclusive shadow, the shorter the string of potential cases. And once again, the scope of that preclusive effect hinges on which procedural rule is adopted, from a range of possible models falling along a familiar spectrum.

Figure 5: Conceptual Approaches to Regulating Serial Prosecutions


At one pole of that spectrum (the leftmost above), a prosecutor unconstrained by the law of preclusion can pursue as many separate cases against a defendant as she likes, stopping only when the marginal value of additional bites at the apple ceases, in her estimation, to be worth her time and effort. Alternatively, at the opposite pole, an initial prosecution for any offense could be deemed to bar further prosecutions of the same defendant for any and all offenses known to the prosecutor when her first case is filed—a “deck-clearing” rule that would effectively eliminate threats of serial prosecution altogether. 63 A broader variation might preclude any offenses the prosecutor should have known about at the time of filing, whether she actually did know about them or not. Cf. W. Va. R. Crim. P. 8(a)(2) (applying a (less sweeping) mandatory joinder rule whenever “two or more offenses are known or should have been known by the exercise of due diligence to the attorney for the state at the time of the commencement of the prosecution”). And between these poles, an initial prosecution’s preclusive effect could turn on the relationship between its underlying facts and the facts of the other cases in the proposed string, with broader or narrower factual relationships creating a secondary spectrum of regulatory approaches.

The law of preclusion thus mirrors the law of joinder: Fact-based regulatory models fall between more absolute extremes, albeit arrayed here in inverse order, given that broad joinder authority enhances prosecutorial power while a broad preclusive shadow restricts it. Crucially, however, the relationship between these two bodies of law also impacts prosecutorial power, as it determines when, if ever, a prosecutor must winnow down the menu of charging options available to her under the substantive law of the jurisdiction. To appreciate this dynamic, note that a given jurisdiction’s definition of joinable offenses can have one of three relationships to the jurisdiction’s law of preclusion: The set of joinable offenses can be broader than, coterminous with, or narrower than the set of precluded offenses—three options that in turn produce a trio of regulatory models, as captured below.

Figure 6: The Relationship Between the Laws of Joinder and Preclusion


In the first of these models (on the left), the universe of joinable offenses is broader than the universe of precluded offenses, which makes joinder permissive: The prosecutor has the option to stack charges together into one large pile or to spread them out into a series of separate piles, without fear of one pile precluding the next. 64 Taking our running example, a permissive regime might give a prosecutor a choice between filing an indictment charging theft, assault, and possession of a weapon all together or instead filing a series of indictments charging those offenses separately. Note, however, that a permissive-joinder regime may still require the prosecutor to join some offenses together in her initial prosecution in order to avoid having those offenses precluded. For example, a state may require prosecutors to join together all charges from the same event but permit prosecutors to choose whether or not to join charges from a common scheme. Here, the regime is “permissive” insofar as the universe of charges that prosecutors are permitted to join is broader than the universe of charges that they must join. In short, the wider the band of joinable-but-not-preclusive offenses, the broader the prosecutor’s discretion will be—up to the theoretical maximum in which joinder is unlimited but preclusion is nonexistent. In the second model, by contrast, joinder and preclusion are coterminous, which makes joinder mandatory: Any offense that can be joined must be joined, or it will be lost to the preclusive shadow of the initial prosecution; but offenses that are not joinable are not subject to preclusion. A prosecutor who wants to save charges from preclusion therefore can—and must—include all joinable charges in her first (and only) bite at the apple. 65 Continuing the example, our hypothetical prosecutor in this regime must charge together all of the robbery offenses that are joinable under the applicable rule, as any that she omits will be precluded; but anything she cannot join in that case (under the applicable joinder rule) she will be free to pursue in another. Finally, in the third model, the universe of joinable offenses is narrower than the preclusive shadow cast by the case in which those offenses are filed. Here, the prosecutor must select at the outset which charges to pursue in that first case and which to forever surrender to the law of preclusion—that is, she must winnow down her charges and focus from the outset on her priorities. 66 Here, our hypothetical prosecutor might be required (by a nonjoinder rule) to pick only one of the robbery-related charges, while at the same time being blocked (by a same-event preclusion rule) from filing any other robbery-related charges in a separate case.

To appreciate the practical significance of these different regimes, consider our hypothetical armed robbery defendant—only now imagine him to be a young man also suspected of engaging, on a separate occasion, in teenage prostitution, under circumstances that depict him (sympathetically) as poor, homeless, and the victim of sexual abuse. 67 Cf. Juan David Romero, Why Do We Treat Child Sex-Trafficking Victims Like Criminals?, New Republic (Dec. 4, 2014),
victims-are-treated-criminals-us [] (“Across much of America, underage sex trafficking victims are treated as criminals rather than victims . . . .”).
A prosecutor attempting to maximize her plea bargaining leverage over such a defendant may want to file two separate piles of charges against him—one for the robbery-related charges and one for the prostitution-related charges. But she may not want to threaten a single omnibus prosecution that combines all these charges together, as she may prefer to insulate the less-sympathetic case from the more-sympathetic one or to hold one in reserve as a looming threat if the first case falters. Moreover, if the two cases involve entirely different witnesses and evidence, litigating them together would be transparently more costly than trying to get a conviction in one of them first. As a result, any added charge-piling leverage associated with an omnibus indictment might be mitigated by the fact that a defendant who senses the prosecutor’s resource constraints could see through her threat to prosecute the robbery and the prostitution offenses at the same time.

Note, however, that the prosecutor’s calculus might change substantially if the robbery defendant’s other suspected crime were not prostitution but rather murder: Now the prosecutor may affirmatively want to pursue an omnibus prosecution (if the law of joinder so permits) on the theory that her odds of conviction are highest if all of the defendant’s alleged misdeeds are put before a single jury. 68 But cf. supra note 50 (discussing Drew v. United States, 331 F.2d 85, 90 (D.C. Cir. 1964)). In short, a prosecutor’s approach to charge piling and case piling is contextually contingent. Sometimes she will want to pile up charges; sometimes she will want to pile up cases; and sometimes she will want to employ a nuanced blend of both tactics. Her ability to pursue any of these approaches, however, depends on the underlying procedural law. A permissive-joinder regime affords her maximum flexibility to choose between or to blend charge-piling and case-piling approaches as she sees fit, whereas a mandatory-joinder regime narrows that flexibility, forcing a prosecutor who prefers to avoid an omnibus prosecution to choose which charges to pursue and which to forgo. 69 Mandatory joinder occupies the middle position on the regulatory spectrum because it does not impose any meaningful constraints in the many cases in which the prosecutor’s incentive is to engage in maximal charge stacking. Indeed, mandatory joinder promotes charge stacking, insofar as it imposes the procedural penalty of preclusion for any charges omitted from the pile. If, however, one assumes that the natural incentives to engage in charge stacking are already sufficient to cause prosecutors to use that tactic as frequently and as robustly as they deem beneficial, then the marginal impact of a mandatory-joinder regime is its imposition of a regulatory constraint in those (rare) cases in which the prosecutor, for strategic or resource-related reasons, prefers to avoid a sprawling omnibus prosecution (as in the robbery–prostitution example above). In these scenarios, prosecutors may indeed stack charges in whichever way is most beneficial to them, but with the added cost that any charges they exclude will be lost to them forever. And a winnowing joinder regime goes a step further, forcing that choice even if the prosecutor actually wants to file the omnibus prosecu­tion—because omnibus prosecutions will generally be prohibited as a matter of law. 70 Note that a winnowing joinder regime might ban only some subset of omnibus prosecutions; it will not necessarily eliminate charge piling altogether if the underlying joinder rule is broader than nonjoinder. For example, if the underlying joinder regime draws its line of joinable offenses at offenses committed during the same factual episode, while the preclusion regime bars subsequent prosecution for any similar offenses, it is conceivable that our hypothetical prosecutor would have to choose between the robbery charges and the prostitution charges—but she would still be able to pile on multiple charges within whichever case she elected. (This example assumes that robbery and prostitution, two financially motivated crimes, would be deemed “similar offenses.” Cf. supra note 47 (discussing definitions of “similar offense”)).

In the face of these multifaceted and intersecting regulatory dynamics, scholarly analysis of serial prosecutions tends to focus on a much narrower question: What does the Double Jeopardy Clause of the federal Constitution prohibit and permit? 71 On the scholarly focus on double jeopardy doctrine, to the exclusion of subconsti­tutional regulation, see Lear, supra note 60, at 626–27 (criticizing the literature’s focus on “constitutional definitions and purposes”). But cf. Daniel C. Richman, Bargaining About Future Jeopardy, 49 Vand. L. Rev. 1181, 1183–84 (1996) (recognizing “that constitutional doctrine may not be the chief determinant of a defendant’s rights” in this arena given institutional constraints); Ryan C. Schotter, Note, State v. Gonzales: Reinvigorating Criminal Joinder in New Mexico, 44 N.M. L. Rev. 467, 468 (2014) (proposing subconstitutional reform). As for pedagogical priorities, a survey of casebooks is revealing. See, e.g., Allen et al., supra note 44, at 1168–83, 1549–1613 (devoting sixteen pages to the subconstitutional law of joinder and severance, and sixty-five pages to the constitutional law of serial prosecutions); Dressler & Thomas, supra note 44, at 912–18, 1308–53 (devoting seven pages to subconstitutional joinder and severance, and forty-six to the constitutional law of serial prosecutions); Kamisar et al., supra note 44, at 1062–68, 1089, 1068–87, 1369–403 (devoting eight pages to subconstitutional joinder and severance, and fifty-five to the constitutional law of serial prosecutions). The limited regulatory significance of that constitutional framework, however, is captured by marking where it falls on our conceptual schematic—reproduced below with the constitutional floor inserted in bold.

Figure 7: Conceptual and Constitutional Approaches to Regulating Serial Prosecutions


As the updated schematic shows, the Double Jeopardy Clause is indeed more restrictive than a pure nonpreclusion regime, insofar as it prohibits serial prosecutions in at least some circumstances. Those circumstances, however, are few and far between: The Clause’s primary implementing doctrine, the same-elements test, bars serial prosecutions only if all of the elements in all of the charges of a proposed second prosecution contain, or are contained within, the elements of a prior offense from the same factual episode for which the defendant was already acquitted or convicted—a test rarely satisfied in a world where criminal codes are full of only partially overlapping offenses. 72 See Brown v. Ohio, 432 U.S. 161, 166 (1977); Blockburger v. United States, 284 U.S. 299, 304 (1932); see also United States v. Perez, 22 U.S. (9 Wheat.) 579, 580 (1824); cf. Stuntz, Politics, supra note 5, at 507, 531 (criticizing “[l]ax double jeopardy doctrine” along these lines). Our hypothetical armed robbery defendant, for example, could likely be prosecuted at least five separate times under this test, as theft, aggravated assault, threats, unlawful possession of a weapon, and brandishing a firearm each typically contains an element that the others do not. And while the same-elements test is supplemented in cases of prior acquittal by the collateral estoppel test, that added layer of protection bars subsequent prosecutions only if a conviction in the second case would require a jury to accept facts that the first jury definitively rejected—a thin reed that “will not often be available,” given that multi-element offenses and general jury verdicts often make it too difficult “to determine with precision” just how the first “jury has decided any particular issue.” 73 5 Wayne R. LaFave et al., Criminal Procedure § 17.4(a), at 64 (4th ed. 2015) (internal quotation marks omitted) (quoting Walter V. Schaefer, Unresolved Issues in the Law of Double Jeopardy: Waller and Ashe, 58 Calif. L. Rev. 391, 394 (1970)); see also Ashe v. Swenson, 397 U.S. 436, 443 (1970) (establishing the collateral estoppel test). By way of example: If our hypothetical armed robbery defendant argued at trial that the prosecutor failed to prove beyond a reasonable doubt that anything of value was actually taken from the victim, or that the perpetrator was armed, or that he was in fact the perpetrator, then the collateral estoppel test would offer him no protection against serial prosecutions even if he is acquitted in his first case, because the words “not guilty” will not reveal which argument swayed the jury.

Legal scholars assessing this constitutional landscape routinely criticize it for offering only the weakest of protections. 74 See, e.g., Richman, supra note 71, at 1188 n.23 (“[Double jeopardy doctrine] has long been attacked as ‘inadequate to provide meaningful protection against multiple prosecutions.’” (quoting George C. Thomas III, The Prohibition of Successive Prosecutions for the Same Offense: In Search of a Definition, 71 Iowa L. Rev. 323, 370 (1986))). And in so doing, they tend to do what criminal procedure scholars do best: propose a better constitutional law. Indeed, much of the scholarly debate centers on whether the Supreme Court should move the current constitutional floor a notch or two to the right on the spectrum, embracing a same-event or same-chain-of-events model as the new rule. 75 See Lear, supra note 60, at 626–27 (noting the long-standing scholarly debate on the “relative merits of a conduct-based compulsory joinder rule . . . centered on the meaning of the ‘same offence’ language of the Double Jeopardy Clause” (footnote omitted)). Notably, however, the Supreme Court has considered precisely such proposals, going so far as to briefly adopt a same-event test in Grady v. Corbin, 76 495 U.S. 508, 521 (1990). only to soundly reject it a mere three years later in United States v. Dixon, 77 509 U.S. 688, 711–12 (1993). which overruled Grady, restored the current baseline, and emphasized the extent to which the Court had “consistently rejected” efforts to instantiate broader constitutional protections. 78 Id. at 709 n.14 (noting the repeated but unsuccessful efforts on the part of Justice Brennan to adopt a “same-transaction” test akin to a chain-of-events test (citing Brown v. Ohio, 432 U.S. 161, 170 (1977) (Brennan, J., concurring))); see also Ashe, 397 U.S. at 449 (Brennan, J., concurring).

Thus, perhaps even more so than with joinder and severance, a constitutionally grounded reform agenda in this arena seems certain to face all but insurmountable doctrinal resistance. And yet, here as before, the constitutional floor is hardly the entire, or even the most significant, component of the governing regulatory framework. For once again, prosecutors’ power to threaten serial prosecutions is regulated in various ways by the subconstitutional procedural law of the states.

Figure 8: Variability Across States in the Law of Preclusion 79 For sources, see infra Appendix at Table A.


As the chart above reflects, a slight majority of states (twenty-seven) stick to the Double Jeopardy Clause’s constitutional baseline. The remainder, however, deploy their subconstitutional procedural law to constrain case piling in various additional ways. Of these, six embrace the same-event approach that the Supreme Court rejected in Dixon; ten embrace the broader chain-of-events approach favored by Justice Brennan but never adopted by his colleagues; 80 See sources cited supra note 78 (offering Justice Brennan’s approach); cf. infra note † (p. 1391) (discussing the hazy definitional line between “event” and “event-chain” models). six more preclude all charges encompassed within the same plan or scheme charged in a prior case; and one state (New Mexico) goes so far as to employ a mandatory similar-offense rule, under which all offenses of the same general nature must be brought together in a single omnibus prosecution, even when such an approach could seriously hinder the prosecutor’s case. 81 See N.M. R. 5-203(A) (“Two or more offenses shall be joined . . . if the offenses . . . are of the same or similar character, even if not part of a single scheme or plan . . . .”); State v. Gallegos, 152 P.3d 828, 833 (N.M. 2007) (“[W]e exercised our supervisory powers in 1979 to change the rule regarding joinder of offenses from permissive to mandatory.”); see also id. at 833–34 (holding that “pursuant to the requirements of Rule 5-203(A), the State appropriately and necessarily charged . . . offenses related to” two separate victims in one case, even though “the evidence pertaining to each victim would not have been cross-admissible” at separate trials). Notably, while this regime forces the prosecutor to choose between pursuing a potentially unwieldy omnibus prosecution or winnowing her charges down ex ante, if she elects the omnibus route, a modified severance-as-of-right rule provides the defendant a second, additional layer of protection by affording him the option to sever out charges that are (in his estimation) too factually disconnected from one another to defend against at once. See id. at 834–42; cf. Leipold, supra note 31, at 1143 (discussing challenges of simultaneously defending against factually disparate charges); supra note 50 (discussing equitable severance and the cross-admissibility-of-evidence rule). Moreover, of the twenty-three states that impose regulations above the constitutional floor, one-fifth do so through mandatory-joinder regimes that constrain prosecutors’ ability to toggle between charge-piling and case-piling tactics.

Finally, while no state currently employs the maximally restrictive “deck-clearing” rule, one state (Montana) came very close, adopting a statute that by its plain terms treated one prosecution as precluding another involving the same defendant even with respect to “unrelated offenses,” so long as the precluded charges were “‘consummated prior to the original charge’” and were “‘known to the prosecutor’” at the time she brought the initial case. 82 See State v. Berger, 856 P.2d 552, 553 (Mont. 1993) (emphasis added) (quoting Mont. Code Ann. § 46-11-503 (2017)). Notably, because the preclusion rule embodied in the statutory text was broader than the state’s joinder rule, it established a de jure winnowing regime in which prosecutors in some cases would be forced to prioritize among the various potential offenses available to them, only a subset of which could legally be filed in the sole prosecution permitted by law. Cf. Mont. Code Ann. § 46-11-404 (adopting similar-offense joinder). The statutory text, however, did contain a “consummated prior” provision, quoted above, that distinguishes it from a true “deck-clearing” rule. Under that provision, a prosecutor could avoid preclusion by proceeding chronologically through the potential cases, such that a defendant suspected of, say, shoplifting one day and aggravated assault the next could force the prosecutor to either pursue the less significant case first or else to abandon it altogether. Initially applied pursuant to its expansive terms by trial courts in the state, that statute was eventually narrowed by the state supreme court—but only via a countertextual statutory interpretation that drew repeated dissents from two justices, each of whom would have embraced the broad “same-defendant” rule embodied in the statutory text. 83 See Berger, 856 P.2d at 555 (Trieweiler, J., dissenting, joined by Hunt, J.) (criticizing the majority for holding “that the Legislature did not mean to say what it clearly did say”); see also State v. Waldrup, 872 P.2d 772, 775 (Mont. 1994) (Trieweiler, J., dissenting, joined by Hunt, J.).

While that broadest of approaches was narrowly rejected in Montana, the very fact that it was captured in the text of a statutory amendment enacted by the legislature and repeatedly endorsed by multiple state supreme court justices—combined with the range of less extreme but still significant regulatory approaches adopted by New Mexico and twenty-one other states—simply underscores the central point: Far more than constitutional doctrine implementing the Double Jeopardy Clause, it is the subconstitutional law of preclusion that regulates prosecutors’ ability to pile cases on top of each other.

C. Sentence Piling: The Law of Cumulative Punishment

In addition to regulating the size and the number of charging piles that a prosecutor can threaten ex ante, procedural law also holds the potential to constrain prosecutorial power on the back end, by defining the circumstances under which multiple charges will yield multiple cumulative sentences or will instead collapse into a single concurrent sentence for the most serious charge in the pile. The issue here is arithmetic: If the respective sentences for theft, armed robbery, and using a firearm during a crime of violence are two, three, and five years, will our hypothetical armed robbery defendant face a potential five-year term (if the sentences are served all together) or a potential ten-year term (if they are served back-to-back)?

As the example shows, limiting the ability to pile on sentences elimi­nates much of the leverage associated with piling on charges. 84 Cf. Richard L. Lippke, The Ethics of Plea Bargaining 34 (2011) (“One way to neutralize the trial penalties that strategic overcharging threatens is to limit the potential sentencing impact of conviction on multiple charges.”); Stuntz, Politics, supra note 5, at 595 n.313 (“[T]he key issue is not what charges the defendant faces, but what consequences flow from those charges. Limit the power to increase the defendant’s sentence, and the manipulation of charges will cease to be attractive.”). Note that a rule requiring concurrent sentences will not eliminate all of the tactical benefits that charge-stacking affords, as a hefty indictment can still increase the likelihood of conviction and can also have psychological effects on the defendant. See supra note 31. And indeed, scholars have noted the “tendency of prosecutors to pad indictments with as many counts as they can derive from the available facts—even when sentences will be concurrent.” Richman, supra note 71, at 1195. Scholars hoping to effectuate such limitations often advocate reforms grounded in familiar constitutional provisions—in this case in the Double Jeopardy Clause, the Eighth Amendment, and the Due Process Clause. 85 See, e.g., Jacqueline E. Ross, Damned Under Many Headings: The Problem of Multiple Punishment, 29 Am. J. Crim. L. 245, 322 (2002) (“Double Jeopardy jurisprudence, if refashioned, could guide trial courts and legislatures in thinking about how to coordinate penalties.”); George C. Thomas III, A Blameworthy Act Approach to the Double Jeopardy Same Offense Problem, 83 Calif. L. Rev. 1027, 1070 (1995) (proposing to “subject” cumulative punishment “to the limits of the Eighth Amendment”); see also Nancy J. King, Portioning Punishment: Constitutional Limits on Successive and Excessive Penalties, 144 U. Pa. L. Rev. 101, 104–05 (1995) (proposing “several modifications to existing Fifth and Eighth Amendment doctrine” to help existing constitutional frameworks “operate in tandem to regulate the totality of punishment”). Indeed, even those inclined to view sentencing law through a subconstitutional lens give subconstitutional law on cumulative sentencing short shrift. See, e.g., Nora V. Demleitner et al., Sentencing Law and Policy 236–43 (2004). But here as before, the constitutional pathways are shut. Longstanding double jeopardy doctrine, for example, makes clear that the Clause does not prohibit cumulative punishments at all but rather offers only a watered-down version of the same-elements test, which operates here as only a default presumption that states are free to ignore—in those rare instances when it applies. 86 See Missouri v. Hunter, 459 U.S. 359, 368–69 (1983) (holding that the same-elements test that on rare occasion bars serial prosecutions “is not a constitutional rule” in the cumulative sentencing context, such that “a legislature [may] specifically authorize[] cumulative punishment under two statutes” that would be the “same offence” under Blockburger v. United States, 284 U.S. 299 (1932)); Seigel & Slobogin, supra note 31, at 1123 (“[T]he Supreme Court has been . . . very clear . . . that legislative intent alone determines the scope of the ‘same offense’ analysis for determining when cumulative punishments may be imposed . . . .”). As for the Eighth Amendment, courts have consistently held “that separate sentences must be considered separately for proportion­ality review purposes,” thus omitting cumulative sentences from the constitutional analysis altogether. 87 Close v. People, 48 P.3d 528, 539 (Colo. 2002); see also id. (observing that “several lower federal courts have similarly held that for purposes of proportionality review each sentence imposed must be considered separately” (citing Pearson v. Ramos, 237 F.3d 881 (7th Cir. 2001); United States v. Aiello, 864 F.2d 257 (2d Cir. 1988); United States v. Schell, 692 F.2d 672 (10th Cir. 1982))). The Supreme Court has shown little inclination to read the Eighth Amendment as protecting against lengthy sentences more generally and has shown an ongoing disinclination to impose constitutional restraints on consecutive sentences in particular. See Ewing v. California, 538 U.S. 11, 17–18, 30–31 (2003) (upholding a twenty-five-year sentence for theft of three golf clubs under a three-strikes law); Harmelin v. Michigan, 501 U.S. 957, 961, 994–95 (1991) (upholding a mandatory life sentence for drug possession); Carol S. Steiker & Jordan M. Steiker, Courting Death 307–08 (2017) (“In the aftermath of these decisions . . . there seemed to be little remaining hope that any sentences of incarceration would be deemed disproportionate under the Eighth Amendment.”); see also Oregon v. Ice, 555 U.S. 160, 169–70 (2009) (exempting consecutive sentences from the holding of Apprendi v. New Jersey, 530 U.S. 466, 490 (2000)). And the Due Process Clause is unlikely to offer protection where these more textually specific clauses do not. 88 See Medina v. California, 505 U.S. 437, 443 (1992) (holding that, “[i]n the field of criminal law . . . ‘the Due Process Clause has limited operation’” because “[t]he Bill of Rights speaks in explicit terms to many aspects of criminal procedure” (quoting Dowling v. United States, 493 U.S. 342, 352 (1990))); Graham v. Connor, 490 U.S. 386, 395 (1989) (holding that when an “[a]mendment provides an explicit textual source of constitutional protection . . . that Amendment, not the more generalized notion of ‘substantive due process,’ must be the guide for analyzing [constitutional] claims”). Thus, as clear-eyed scholars urging constitutional reform readily acknowledge, their proposals rest on “weak doctrinal pedigree” and would ultimately require a “radical . . . change in constitutional law.” 89 Stuntz, Politics, supra note 5, at 594–96 (“There is . . . no line of cases that lays the doctrinal foundation for constitutionalizing judges’ opportunity to show mercy to those defendants who, in the judges’ eyes, deserve it.”); id. at 600 (promoting a constitutional reform agenda that “sounds radical” because “in some ways it is”); see also King, supra note 85, at 153–54, 184 (acknowledging that the “Eighth Amendment is [currently] of little use as a limit on disproportionate punishment that results from cumulative penalties” but arguing that “[a] more inclusive grouping formula is needed”).

Not so, however, for subconstitutional law, which is home to the legal frameworks that actually establish each state’s cumulative-sentencing regime in the first instance. To appreciate how such frameworks might constrain prosecutorial power, consider four basic models that capture the various regulatory options available.

Figure 9: Conceptual Approaches to Regulating Multiple-Offense Sentencing


The most empowering model for prosecutors (the leftmost above) simply requires consecutive sentences as a matter of law, thus guaranteeing that every additional charge a prosecutor files will produce a longer sentence if the defendant is convicted. 90 Cf. Meares, supra note 17, at 888 (noting that prosecutorial control over charges translates into control over “punishment if the sentences for the various charges are imposed consecutively rather than concurrently”). A variant on mandatory-consecutive sentencing could require the court to impose “additional punishment” for each additional charge of conviction, “but in progressively diminishing amounts.” Stephen Breyer, The Federal Sentencing Guidelines and the Key Compromises upon Which They Rest, 17 Hofstra L. Rev. 1, 25–28 (1988). Alternatively, at the opposite pole, the law can mandate concurrent sentences, thereby eliminating prosecutors’ ability to increase defendants’ sentencing exposure unilaterally. 91 As noted supra note 84, a mandatory-concurrent regime would not eliminate all of the tactical advantages that charge piling affords. And between these poles, various discretionary approaches could check sentence piling without extinguishing cumulative sentences entirely. Thus, under an open-discretion model, judges would be allowed to sentence consecu­tively or concurrently as they see fit—perhaps nudged toward one option or the other by default rules favoring consecutive or concurrent sentences. 92 On the power of default rules and nudges, see generally Richard H. Thaler & Cass R. Sunstein, Nudge: Improving Decisions About Health, Wealth & Happiness 1–102 (2009). In such a regime, the judge could check the prosecutors’ charge stacking on the back end, but—importantly—only to a degree: Because sentencing does not take place until after plea negotiations are over, a defendant assessing a prosecutor’s pile of charges ex ante won’t know whether the judge is going to whittle the pile down with a concurrent sentence after the deal is struck—an uncertainty that safeguards prosecutors’ leverage even in a discretionary regime. 93 Cf. Stuntz, Politics, supra note 5, at 519–20 (“By threatening [multiple] charges, prosecutors can, even in discretionary sentencing systems, significantly raise the defendant’s maximum sentence, and often raise the minimum sentence as well. The higher threatened sentence can then be used as a bargaining chip, an inducement to plead guilty.”). To be sure, in individual cases, defendants might be able to glean ex ante signals about how a judge might sentence, particularly if they know who the sentencing judge will be and know that judge’s reputation. Alternatively, in the absence of strong reputational indicators, the subconstitutional procedural law might permit a judge to tell the defendant how she intends to sentence before the plea deal is finalized—or it might prohibit such communications. See infra note 198 (discussing judicial participation in plea bargaining). Assuming reliable signals of judges’ intentions are available, a discretionary regime will constrain prosecutorial power precisely to the extent that judges use their discretion to do so. Cf. supra note 51 (discussing a similar dynamic with equitable severance); infra section V.B (discussing judicial checks on prosecutorial power more generally). That uncertainty can be alleviated, however, by a final model, in which codified factors structure judges’ discretion, thus clarifying in advance those cases in which concurrent sentences are most likely and in which charge piling will therefore be least effective. 94 Note that a structured approach is not inexorably more constraining than an open-ended approach: Because a multifactor test will, by definition, not require concurrent sentences in all cases, it will necessarily be less constraining than a fully discretionary test exercised by a judge who always opts for concurrent sentences. Moreover, the constraint associated with a multifactor approach depends on what the factors are.

With these four ideal types in mind, consider now the various approaches to cumulative sentencing actually employed across the states, beginning with each state’s generic sentencing framework—that is to say, the framework that applies in the absence of some more specific statute carving out an exception.

Figure 10: Variability Across States in the Law of Cumulative Sentencing 95 For sources, see infra Appendix at Table B.


As the chart above shows, the vast majority of states (thirty-seven) employ an open-ended discretionary regime, either in a neutral fashion (as in seventeen states) or with default rules nudging toward consecutive sentencing (in five states) or toward concurrent sentencing (in fifteen states). Thirteen states, however, depart from this open-ended discretionary model. Ten of those offer a list of factors that clarify in advance when concurrent and consecutive sentences should be imposed. And an addi­tional three (Colorado, Michigan, and Texas) go so far as to require concurrent sentences as a matter of law.

These generic frameworks, however, tell only part of the story, for virtually every state supplements its generic regime with one or more subsidiary provisions that carve out exceptions to the general rule, typically by imposing one or more mandatory rules in specific circumstances.

Figure 11: The Subsidiary Law of Cumulative Sentencing 96 For sources, see infra Appendix at Table B.


Thus, forty-two states require consecutive sentences in some specified subset of cases—for example, when certain crimes are at issue or when the defendant is a recidivist who committed a second offense while under supervision for a prior one. At the same time, eighteen states require concurrent sentences in certain circumstances—usually by imposing a cap on the aggregate length of consecutive sentences (and thus requiring concurrent sentences once that cap is reached). 97 Most frequently, this approach is employed when the convictions at issue are misdemeanors. See infra Appendix at Table B. Concurrent sentencing mandates can also be implemented via merger rules, that is, rules that permit imposition of a conviction on only a subset of charged offenses. Such merger rules typically either reaffirm or expand upon the constitutional presumption against imposing multiple sentences for offenses with the same elements. See supra note 86 and accompanying text (discussing the constitutional presumption); see, e.g., Ga. Code. Ann. § 16-1-7(a) (2011) (providing that a defendant “may not . . . be convicted of more than one crime if . . . [o]ne crime is included in the other; or . . . [if] crimes differ only in that one is defined to prohibit a designated kind of conduct generally and the other to prohibit a specific instance of such conduct”); see also id. § 16-1-6 (defining an “included” offense in broader terms than the same-elements test).

In short, given variability in both the baseline sentencing frameworks and the ways in which exceptions to those baselines can be crafted, there are as many different ways of building a cumulative-sentencing regime as there are states themselves. And within that wide range, the regimes actually employed on the ground run the gamut, with states enacting mandatory-consecutive regimes, mandatory-concurrent regimes, and a range of discretionary models in between. Considering the dearth of constitutional regulation in this arena, it is these varied legal frameworks—not some hoped-for constitutional rule—that define the scope of prosecutors’ power to pile on cumulative sentences.

III. Overreaching

Beyond inflating the size, number, or weight of the piles of charges that a defendant could face, a prosecutor can also inflate the individual charges themselves, in one of three ways: First, she can overreach beyond the facts of the case, filing charges that are unlikely to satisfy the burden of proof at trial. Second, she can overreach on the law, filing charges that rest on aggressive interpretations of the statutes or precedents defining criminal liability. Finally, she can overreach beyond the equities of the case, filing charges that are simply too severe, even if the facts and the law could technically support a conviction.

In each instance, the prosecutor trades on the defendant’s fear that a jury might go along with the inflated charges—fear that the prosecutor can exacerbate through even greater overreaching, and that the defendant can extinguish only by pleading guilty to some lower offense. Therein, however, lies the rub, for if the defendant pleads guilty there will not be any adjudication on the merits, and one of the main checks against overreaching will disappear. Regulating overreach thus requires interposing some other check, beyond a trial, on the factual, legal, and equitable merit of the charges. And by determining whether those checks will exist and the form they might take, the subconstitutional law of criminal procedure regulates another significant component of prosecutors’ plea bargaining power.

A. Factual Overreaching: The Law of Pretrial Evidentiary Review

To make factual overreaching concrete, consider once again our hypothetical armed robbery defendant, only imagine now that the sole witness against him is a blind convicted drug dealer with an axe to grind against the defendant. A defense attorney reading that sentence will rightly conclude that the prosecutor has a weak case, with significant—but, important­ly, still uncertain—odds of ending in an acquittal. 98 The uncertainty arises from the fact that the jury could believe the witness—who might in fact be telling the truth. Indeed, as defendants and their attorneys know well, the risk of conviction in even the weakest-seeming case is almost always too high to dismiss out of hand, given the consequences at stake. Cf. Geoffrey A. Campbell, In the Shoes of the Wrongly Accused, A.B.A. J., June 1995, at 32, 32 (reporting that nearly one-third of defendants lack confidence that “a jury would reach a fair verdict if they were accused of a crime they did not commit”); Alanna Durkin Richer & Curt Anderson, Trial or Deal? Some Driven to Plead Guilty, Later Exonerated, Associated Press (Nov. 15, 2016),
24cfa961d3444be49901496fdcaa3fda [] (reporting roughly 1,600 exonerations of individuals wrongly convicted at trial).
If, however, the prosecutor believes her witness, or has some other reason to pursue criminal charges against the defendant, she is free to prosecute him—and to exercise whatever attendant leverage she can to induce him to plead guilty, the weakness of her case notwithstanding. 99 Prosecutors, of course, hardly ever set out to convict innocent defendants and in fact often will not proceed if they are not personally confident of the defendant’s guilt. Cf. Berger v. United States, 295 U.S. 78, 88 (1935) (“[The government’s interest] in a criminal prosecution is not that it shall win a case, but that justice shall be done.”). However, as numerous scholars observe, “various pressures on prosecutors . . . can cause them to act in ways that subvert justice, [either] intentionally or, as is more often the case, unintentionally.” Keith A. Findley & Michael S. Scott, The Multiple Dimensions of Tunnel Vision in Criminal Cases, 2006 Wis. L. Rev. 291, 295 (summarizing “[c]onsiderable literature” recounting such pressures and arguing that “the problem is more pervasive than even that literature suggests”). Perhaps most notably, prosecutors can suffer from “tunnel vision” that blinds them to the weaknesses in their cases, convincing them of a defendant’s guilt even when a more detached observer might harbor serious doubts. See id. at 292–95; see also Daniel S. Medwed, Emotionally Charged: The Prosecutorial Charging Decision and the Innocence Revolution, 31 Cardozo L. Rev. 2187, 2201–05 (2010) (describing how prosecutorial tunnel vision can produce false convictions). Moreover, prosecutors sometimes have an incentive to pursue charges that they know are weak, perhaps to induce the defendant to cooperate with another investigation or to punish him for suspected misconduct that, for various reasons, is simply too difficult to prove in court. Cf. United States v. Kupa, 976 F. Supp. 2d 417, 420 (E.D.N.Y. 2013) (“[T]o coerce cooperation . . . prosecutors routinely threaten ultra-harsh, enhanced mandatory sentences that no one—not even the prosecutors themselves—thinks are appropriate.”); United States v. Dossie, 851 F. Supp. 2d 478, 487–88 (E.D.N.Y. 2012) (discussing cooperation agreements); Daniel C. Richman & William J. Stuntz, Al Capone’s Revenge: An Essay on the Political Economy of Pretextual Prosecution, 105 Colum. L. Rev. 583, 588–99 (2005) (discussing the dynamics that produce pretextual prosecutions).

Under traditional conceptions of the criminal justice system, the pri­mary protection against such overreaching is the defendant’s constitutional right to a trial, the “basic purpose” of which “is the determination of truth.” 100 Tehan v. United States ex rel. Shott, 382 U.S. 406, 416 (1966); see also Duncan v. Louisiana, 391 U.S. 145, 149 (1968) (establishing the right to a jury trial). Indeed, much of criminal procedure’s constitutional edifice is designed to ensure the robustness of that protective device: The defendant enjoys a constitutional right to competent representation at government expense, a constitutional right to present and cross-examine witnesses, and, perhaps most notably, a constitutional right not to be convicted unless the prosecutor convinces a jury that he is guilty beyond a reasonable doubt—that is to say, convinces a jury that she has not overreached. 101 See Crawford v. Washington, 541 U.S. 36, 59 (2004) (describing the right to cross-examine witnesses); Strickland v. Washington, 466 U.S. 668, 669 (1984) (establishing the right to effective counsel); Davis v. Alaska, 415 U.S. 308, 315–18 (1974) (describing the right to present evidence); In re Winship, 397 U.S. 358, 364 (1970) (requiring proof of guilt beyond a reasonable doubt); Gideon v. Wainwright, 372 U.S. 335, 335 (1963) (establishing the right to appointed counsel for indigent defendants). All of those protections, however, disappear if the prosecutor successfully induces the defendant to waive his trial and plead guilty. Indeed, avoiding those costly procedural checks is often the prosecutor’s prime objective. 102 Cf. Frank H. Easterbrook, Criminal Procedure as a Market System, 12 J. Legal Stud. 289, 295–96 (1983) (modeling prosecutors as individuals who “attempt[] to obtain the maximum deterrence from [their] available resources”). And in the course of building pressure to induce such a plea, she will often have a serious “temptation to file more serious charges than the evidence can support,” 103 Wright & Miller, supra note 10, at 85; see also Meares, supra note 17, at 869 (“By charging the defendant with the most serious offenses [available], the prosecutor can push up the trial penalty and limit, as a consequence, the defendant’s ability to waive his right to trial . . . voluntarily.”); id. at 865 (describing how prosecutors exploit the gap “between the quantity and quality of evidence necessary to support a legitimate charge and the quantity and quality of evidence needed to prove that the defendant committed the charged offense”). To be sure, if the prosecutor aims too high, she could overplay her hand and thereby risk an acquittal, which would undermine her charge-bargaining leverage. Such a risk, however, is substantially mitigated if the prosecutor enjoys wide latitude to “fall back” from an overplayed hand to a less aggressive one—a dynamic discussed infra Part IV (examining “sliding down”). a tactic that when combined with the charge piling discussed in Part II has rendered criminal trials all but extinct. 104 See infra Table 2 (p. 1375) (reporting statewide trial rates ranging from 1.5% to 5.5%).

In this trial-free world, any procedural check on factual overreaching must come not from juries but rather from the only other neutral actors at hand: judges, exercising some form of pretrial judicial review. 105 “Judge” here is meant broadly to include magistrates and commissioners. While one might also imagine grand juries as a check on prosecutorial overreach, the grand jury’s “reputed ability to protect [against] unfounded prosecution” has long been discredited, to the point that “few scholars or practitioners take its screening function seriously today.” Peter Arenella, Reforming the Federal Grand Jury and the State Preliminary Hearing to Prevent Conviction Without Adjudication, 78 Mich. L. Rev. 463, 474 (1980); see also Ortman, supra note 1, at 516 n.16 (“A substantial body of scholarship identifies reasons . . . why grand juries have been rendered irrelevant in modern criminal justice.”). But cf. Ric Simmons, Re-Examining the Grand Jury: Is There Room for Democracy in the Criminal Justice System?, 82 B.U. L. Rev. 1, 37–39 (2002) (defending grand juries). And yet, the mechanisms of such review go almost entirely unexamined in criminal procedure literature, which often assumes that pretrial judicial review of prosecutors’ charging decisions either does not or cannot exist. 106 See, e.g., Carrie Leonetti, When the Emperor Has No Clothes: A Proposal for Defensive Summary Judgment in Criminal Cases, 84 S. Cal. L. Rev. 661, 691 (2011) (calling the “lack[] [of] an official remedy for courts to employ when the prosecution charges an offense that it has no chance of proving” one of the “primary problems with the current system”); Wright & Miller, supra note 10, at 112–13 (noting the scholarly view that “separation of powers” principles tend to render “limits on charging . . . impossible for other institutions to enforce,” and arguing that “controlling charge bargains depends on prosecutors to regulate themselves in an area where other institutions traditionally take no action”); see also Ortman, supra note 1, at 513 (observing that “criminal procedure literature has paid surprisingly little attention to” the “[c]harging standards [that] are essential elements of criminal justice systems”). The most significant discussion of pretrial charge review in the literature to date is Professor William Ortman’s article urging abandonment of the probable cause standard employed by federal grand juries. See id. at 519–40, 568. Ortman, however, limits his analysis to the grand jury “charging standard itself” as it is employed in federal courts, bracketing both “the institutional mechanism[s] for enforcing” that standard, id. at 516 n.16, and also other modes of pretrial screening beyond the grand jury process, see id. at 546 n.201. Apart from Ortman’s work, the most substantial treatment of pretrial judicial screening in criminal cases is an article written nearly four decades ago that also focuses heavily on grand juries and federal processes. See Arenella, supra note 105, at 481–96. For a sampling of other early work, see 4 LaFave et al., supra note 73, § 14.1(a), at 309 n.6 (citing sources and observing that, “[u]nfortunately, all of [them] are somewhat dated”). That assumption is accurate as far as constitutional law goes, given longstanding precedent holding that there is no constitutional right to pretrial judicial review of the evidence supporting “the formal accusation [filed] by the district attorney.” 107 Lem Woon v. Oregon, 229 U.S. 586, 590 (1913); see also Barkow, Separation of Powers, supra note 11, at 1044 (“The Court allows prosecutors almost unlimited discretion to make charging decisions.”); cf. Armstrong v. United States, 517 U.S. 456, 464 (1996) (calling the decision to file a criminal charge “a ‘special province’ of the Executive” that “‘courts presume’ . . . ‘rests entirely in [the prosecutor’s] discretion.’” (first quoting Heckler v. Chaney, 470 U.S. 821, 832 (1985); then quoting United States v. Chem. Found., Inc., 272 U.S. 1, 14–15 (1926); and then quoting Bordenkircher v. Hayes, 434 U.S. 357, 364 (1978))). The closest constitutional law comes to requiring a form of pretrial judicial review is its requirement that a judge review the facts underlying a defendant’s warrantless arrest—a form of review directed at police officers, not prosecutors, and that does not bar prosecutors from filing charges even if a judge deems the associated arrest invalid. See Gerstein v. Pugh, 420 U.S. 103, 113–14, 116–19 (1975). Pretrial evidentiary review, however, is in fact a commonplace of the American criminal process—and it is executed in strikingly different ways across jurisdictions, albeit always through mechanisms established by subconstitutional procedural law.

To unpack this complex set of legal frameworks, it is helpful to identify at the outset three core questions that this body of law resolves: First, will the factual basis of the prosecutor’s charges be tested via an adversarial evidentiary hearing? 108 The hearing is often called a “preliminary hearing,” but other names include “the ‘probable cause’ hearing, the ‘commitment hearing,’ the ‘examining trial,’ and the ‘bindover’ hearing.” 4 LaFave et al., supra note 73, § 14.1(a), at 307–08. Second, in the absence of such a hearing, will the evidence supporting the charges be exposed to any other form of judicial review? Finally, if there is an evidentiary hearing, how robust will it be in probing the factual allegations?

As these questions indicate, pretrial evidentiary hearings are the lynchpin of this regulatory regime, for they offer a potentially powerful check on prosecutorial overreach: cross-examination of the prosecutor’s witnesses. 109 While the robustness of the hearing can vary in important ways, see infra text accompanying notes 133–144, “[a]ll jurisdictions” that have preliminary hearings “grant the defense a right to cross-examine those witnesses presented by the prosecution.” 4 LaFave et al., supra note 73, § 14.4(c), at 392; see also Coleman v. Alabama, 399 U.S. 1, 9–10 (1970) (guaranteeing the right to counsel at a preliminary hearing, if the hearing exists); cf. California v. Green, 399 U.S. 149, 158 (1970) (describing cross-examination as “the ‘greatest legal engine ever invented for the discovery of truth.’” (quoting 5 J. Wigmore, Evidence 1367 (3d ed. 1940))). In order to appreciate the range of different procedural models governing these hearings, however, it is important to recognize up front a curious feature of this particular landscape: Oftentimes, the prosecutor herself can decide whether to subject herself to such a hearing or whether instead to bypass it altogether. The historical origins of this bypass authority are tied to the grand jury, the idea being that if that putatively neutral and independent body reviews the factual allegations at hand and returns an indictment, further review by a judge would be superfluous. 110 See, e.g., Sciortino v. Zampano, 385 F.2d 132, 133 (2d Cir. 1967) (“[T]he return of an indictment . . . establishes probable cause [and] eliminates the need for a preliminary examination . . . [which] would be an empty ritual . . . .”). Of course, in the modern era, grand juries are widely understood to return indictments whenever a prosecutor asks them to, which means that the (anachronistic) power to bypass a preliminary hearing simply by obtaining an indictment is now a power that for all practical purposes is enjoyed by the prosecutor herself. 111 On the rubber-stamp nature of grand juries, see supra note 105; see also Editorial, Do We Need Grand Juries?, N.Y. Times (Feb. 18, 1985),
opinion/do-we-need-grand-juries.html (on file with the Columbia Law Review) (attributing to Judge Sol Wachtler the aphorism that a grand jury “would indict a ham sandwich”). Indeed, far from imposing a substantive check on prosecutors’ charging authority, grand juries often afford prosecutors important “investigative advantages,” such as pretrial subpoena power. 3 LaFave et al., supra note 73, § 8.3, at 23. For an example of anachronistically path-dependent reasoning offered in support of indictment-bypass authority, see Kaley v. United States, 134 S. Ct. 1090, 1097–98 (2014) (citing the “historic commitment of our criminal justice system” as reason to reaffirm the principle that the “grand jury gets to say—without any review, oversight, or second-guessing—whether probable cause exists to think that a person committed a crime”).

Note, however, that while grand juries impose virtually no substantive constraints on prosecutors’ charging decisions, the process of assembling the grand jury and presenting witnesses to it does take time and energy. 112 See Crane, supra note 23, at 803–04 (noting that even if “grand juries rarely decline to indict, the grand jury requirement still imposes meaningful costs . . . related to prosecutor time and grand jury time”). Thus, if the time frame within which that process must unfold is relatively short, if the number of cases in need of processing is high, and if prosecu­torial resources are low, an indictment-bypass rule will constrain prosecutors to a degree: They will be forced to choose which cases to send to the grand jury and which to (reluctantly) allow forward to a preliminary hearing. 113 Notably, the time frame within which an indictment must be returned in order to trigger a bypass is itself decided by the subconstitutional procedural law, with timelines ranging between two and sixty days after a defendant’s arrest (depending in part on whether the defendant is detained prior to trial). Compare Ariz. R. Crim. P. 5.1(a) (requiring a preliminary hearing “no later than 10 days after the defendant’s initial appearance if the defendant is in custody, or no later than 20 days after the defendant’s initial appearance if the defendant is not in custody”), with Haw. R. Penal P. 5(c)(3) (mandating preliminary hearings within two days for defendants in custody, and thirty days for defendants not in custody), and N.M. R. 6-202 (mandating preliminary hearings within ten days for defendants in custody, and sixty days for defendants not in custody). Moreover, if “prosecutors intend to bypass but cannot obtain an indictment with sufficient promptness, they may” in some jurisdictions “obtain continuances . . . [of] the preliminary hearing” to buy more time. 4 LaFave et al., supra note 73, § 14.2(c), at 338. Crucially, however, some jurisdictions detach the prosecutor’s power to bypass a preliminary hearing from its historical moorings altogether, allowing her to exercise that power not just by filing an indictment but by filing an information as well—that is to say, by writing her chosen charges down on a piece of paper and signing her name. 114 In about one-third of states (and the federal system), felony charges must be filed by indictment, which means that the grand jury process is essentially mandatory. See 4 LaFave et al., supra note 73, § 15.1(d), at 468 (listing eighteen indictment states); id. § 15.1(e), at 478 (noting four states that require indictments only in the most serious cases). However, in roughly two-thirds of the states, prosecutors have the option to file formal charges by either indictment or by information—and almost always choose the latter route, which is far less resource-intensive. See id. § 14.2(d), at 341; see also id. § 15.1(g), at 491–93 (listing states where charging by information is an option and noting that while practice “varies tremendously” across localities, “the grand jury is not part of the standard practice” in “a vast majority” of those states); Crane, supra note 23, at 804 (“[I]f an office has thousands of ham sandwiches to process, the more that can be charged by information the better.”). Here, because the prosecutor can easily and unilaterally preempt the hearing, it will only ever take place in those rare cases in which she wants it to. 115 See 4 LaFave et al., supra note 73, § 14.2(c), at 338–39 (listing “special circumstances” that might render a preliminary hearing “desirable from the prosecutor’s perspective,” including “the need to [preserve] the testimony of a witness” whom the prosecutor expects to be unavailable or unreliable at trial). Alternatively, a prosecutor might on rare occasion opt for a preliminary hearing in order to demonstrate the strength of her case to a defendant whom she believes is refusing a plea deal due to irrational overconfidence. Cf. State v. Blanchard, 786 So. 2d 701, 706 (La. 2001) (observing that “offering [the defendant] a preview” of particularly strong evidence against him may have the “inevitable . . . and permissible” effect of exerting “pressure on [him] to reconsider his decision to stand trial”).

Taking indictment bypass and information bypass together, we can now identify five basic procedural models governing the availability of pretrial evidentiary hearings, which once again can be arrayed along a spectrum.

Figure 12: The Law of Preliminary-Hearing Bypass


Note that the x axis of our otherwise-familiar spectrum is expanded vertically here in order to accommodate the fact that some of these models overlap with one another when it comes to the degree to which they constrain prosecutorial power. 116 There is, in other words, no y axis in this figure and thus no significance to the models’ relative vertical placement. That overlap is minimized at the extremes: At one pole (the leftmost above), the information-bypass model just described maximizes prosecutorial power by giving prosecutors the ability to control whether a preliminary hearing will take place, and to exercise that control with no more effort than the stroke of a pen. At the opposite pole, by contrast, the defendant has an absolute right to a prelimi­nary hearing, meaning there will be one if he wants it. 117 But cf. infra note 211 (discussing the possibility that the defendant will waive his right to such a hearing in exchange for plea bargaining concessions).

As for the more complicated models in the middle of the schematic, the easiest to plot is one in which preliminary hearings simply do not exist—a model that strongly favors prosecutors, who typically want to avoid adversarial testing of their evidence prior to trial. Indeed, compared to an information-bypass model, the no-hearing model disadvantages prosecutors only in those rare cases in which they might actually want the hearing to take place. 118 See supra note 115.

By contrast, the indictment-bypass model is harder to plot, because (as noted above) the constraint it imposes on prosecutors depends on their resource capacity: If the prosecutor can speed her cases through the grand jury, an indictment-bypass model will be almost as favorable to her as an information-bypass model, as it will give her control over whether the hearing takes place while imposing only a modest administrative burden. But if resources are tight, the indictment-bypass model will start to impose genuine constraint, forcing the prosecutor to submit to preliminary hearings in at least some cases when she would rather not. 119 In practice, these resource constraints (combined with the varying time frames within which indictment-bypass must be completed, see supra note 113) yield different uses of the indictment-bypass device across jurisdictions. See 4 LaFave et al., supra note 73, § 14.2(c), at 338–39 (observing that bypass authority is used “regularly,” but nonetheless to different extents across jurisdictions where it exists); cf. id. (observing that “the ready availability of an indicting grand jury leads to more frequent use of the bypass procedure”); id. § 14.2(b), at 334–35 (noting that indictment-bypass authority has “largely . . . eliminate[d] preliminary hearings in the federal courts”).

Finally, the last model in the schematic gives judges discretion to hold preliminary hearings regardless of whether an indictment or information is filed. Like the indictment-bypass model, this model spreads across much of the spectrum, because some judges may opt to hold hearings as a matter of course while others may rarely hold them at all. 120 Cf. infra section V.B (discussing judicial incentives vis-à-vis prosecutorial oversight).

With these different models of prosecutorial bypass now in place, our conceptual schematic is almost complete. The one remaining layer of complexity stems from the fact that, in some jurisdictions, judges can employ a less robust form of evidentiary review in place of or in addition to a preliminary hearing: Instead of examining live witness testimony, the judge might be empowered to review only documentary evidence submitted by the prosecutor in support of her charges—a form of “papers review” that could examine, for example, written statements or grand jury testimony from some selection of witnesses. 121 To the extent that a preliminary hearing can sometimes resemble a “mini-trial,” see 4 LaFave et al., supra note 73, § 14.3(b), at 368–70, papers review might be analogized to a motion for summary judgment, see id. § 14.2(d), at 344–46. The analogy, though, has its limits: Because pretrial discovery in the criminal system is often anemic, with defendants enjoying minimal access to basic document production, let alone depositions or interrogatories, the “papers” being reviewed will often be thin—and a far cry from the reams of evidence underlying many civil summary judgment motions. Cf. Ben Grunwald, The Fragile Promise of Open-File Discovery, 49 Conn. L. Rev. 771, 779–81 (2017) (recounting the “well-known” “limitations of the traditional approach to discovery” in criminal cases). That said, papers review will generally be more robust than the constitutionally required review of the defendant’s arrest affidavit, which occurs within forty-eight hours of the defendant’s arrest at a time when he may not even have an attorney and when the prosecutor herself may not know much beyond what is in the initial police report. See County of Riverside v. McLaughlin, 500 U.S. 44, 56–57 (1991) (requiring Gerstein review “within 48 hours of arrest,” absent “a bona fide emergency or other extraordinary circumstance”); see also 3 LaFave et al., supra note 73, § 11.2(b), at 702–04 (observing that Rothgery v. Gillespie County, 554 U.S. 191 (2008), “strongly suggest[s]” that the right to counsel does not apply at a Gerstein hearing); cf. supra note 107 (distinguishing Gerstein review from pretrial charge review).

Generally speaking, this papers review presents a binary proposition: It exists, or it doesn’t. 122 In truth, the treatment of papers review as a binary proposition oversimplifies matters, as the manner of executing such review can and does vary—most notably with respect to whether or not the review is ex parte. For the sake of simplicity, however, the conceptual schematic above elides these nuances, leaving them to the intrepid reader of footnotes and appendices. See infra notes 126–127; infra Appendix at Table C (citing sources). And by incorporating it into the schematic above, we gain three additional procedural models that, when combined with the initial five, fill in the conceptual map of pretrial evidentiary review. These new models are highlighted in dark gray in the updated schematic below.

Figure 13: Conceptual Frameworks Governing Pretrial Evidentiary Charge Review


The first new model, marked in the third numbered position above, sup­ple­ments the otherwise-maximally-empowering information-bypass regime with papers review, thus ensuring at least some degree of judicial review in all cases. Close on that model’s heels is a regime in which papers review stands alone as the sole mode of pretrial evidentiary review—a model identical to its predecessor save for the fact that it denies prosecutors the option of a preliminary hearing if they want one. 123 See supra note 115. And finally, the third new model layers papers review onto the indictment-bypass regime, authorizing judges to review the grand jury record to ensure indict­ments are factually supported. 124 See generally Sara Sun Beale et al., Grand Jury Law and Practice § 9:26 (2d ed. 2016) (discussing judicial review of the grand jury record).
Notably, nearly all of these eight models are represented among the procedural frameworks actually employed across the states.

Figure 14: Variability Across States in the Law of Pretrial Judicial Charge Review 125 For sources, see infra Appendix at Table C.


Thus, as the chart above reflects, the majority of states (thirty-three) employ an indictment-bypass regime—although more than one third of those (twelve) supplement that prosecutor-friendly regime with papers review of the grand jury record, thereby guaranteeing a judicial check even when prosecutors have the resources to evade preliminary hearings via the grand jury process. 126 The robustness of this supplemental layer of papers review varies: Some states narrowly permit dismissal of an indictment only if the record shows it was supported by zero evidence. See, e.g., State v. Burkhart, 615 S.W.2d 565, 568 (Mo. Ct. App. 1981); Sheriff v. Miley, 663 P.2d 343, 344 (Nev. 1983); State v. Turner, 150 S.E.2d 406, 410–11 (N.C. 1966). Others allow dismissal only if the entirety of the evidence before the grand jury was illegally obtained. See, e.g., Ala. R. Crim. P. 12.8(f)(2). Yet others are more stringent in their review, calling for dismissal if the evidence submitted to the grand jury would not “warrant a conviction” at trial. Alaska R. Crim. P. 6(q); Nelson v. State, 628 P.2d 884, 889 (Alaska 1981). In an additional five states, however, the prosecutor can bypass such a hearing regardless of her resource constraints, simply by filing an information—although here too each of these states employs a form of papers review as a backstop. 127 The extent of papers review in these states varies. In one, the court reviews the information ex parte before it is filed to determine, based on a digest summarizing the prosecutor’s anticipated evidence, whether information submitted “would warrant a conviction by the trial jury.” Iowa Ct. R. 2.5(4). Two others conduct similar review on the defendant’s motion after the information is filed. See Fla. R. Crim. P. 3.190(c)(4) (requiring dismissal when “the undisputed facts do not establish a prima facie case of guilt”); State v. Paredes, 191 So. 3d 936, 939–40 (Fla. Dist. Ct. App. 2016) (equating the prima facie standard to evidence sufficient to survive a motion for summary judgment in a civil case); see also State v. Knapstad, 729 P.2d 48, 51–55 (Wash. 1986) (adopting the Florida approach). Additionally, two of these states grant the defendant a right to call witnesses to supplement the papers-review hearing, while limiting the prosecution to its documentary evidence absent leave from the court to supplement that record. See Haw. Rev. Stat. Ann. § 806-87 (LexisNexis 2016); 12 R.I. Gen. Laws § 12-12-1.8 (2002). By contrast, two other states deem the review satisfied so long as the prosecutor certifies that there are disputed factual questions at issue in the case. See Fla. R. Crim. P. 3.190(c)(4), (d); Knapstad, 729 P.2d at 54. Another two states rely on papers review as their sole method of oversight, 128 One of these states affords a limited right to a preliminary hearing but only in extreme cases carrying a sentence of death or life in prison. See Conn. Gen. Stat. Ann. § 54-46a(a) (West Supp. 2017). Additionally, while rules in these states may technically permit a defendant to supplement papers review through a corresponding evidentiary hearing, such hearings appear to be exceptionally rare. See Vt. R. Crim. P. 12(d)(2) (“The court may determine the motion without hearing as long as both the state and defense have been provided an opportunity to submit affidavits, depositions, or other sworn admissible evidence in written or recorded form.”); id. reporter’s note to 2008 amendment (noting that the rule was “amended to eliminate the necessity of having to schedule a formal hearing on defendant’s motion” and to encourage instead disposition “on the pleadings and written submissions”); see also State v. Florence, 239 N.W.2d 892, 900 (Minn. 1976) (noting that “the presiding judge will [typically] make the critical determination based on all the information contained in the files and records” and that it will be “the rare case where the defendant supports his . . . motion by the production of witnesses”). Accordingly, these states are distinguished from those that have regularly functioning preliminary hearings insulated from prosecutorial bypass by judicial discretion (that is, from model six in Figure 13). while three more deny evidentiary review of prosecutors’ charges at all, 129 In each of these states, the Fourth Amendment’s required Gerstein review of the facts underlying the defendant’s arrest is the only form of pretrial judicial review provided. See Ark. R. Crim. P. 8.3(c) (requiring only “an informal, non-adversary hearing” at the defendant’s initial appearance to determine “whether there is probable cause for detaining the arrested person pending further proceedings”); see also Arkansas Criminal Code Revision Commission’s Bill, Act 1994, § 502, 2005 Ark. Acts 6932, 7450–59 (repealing preexisting preliminary-examination provisions of state law); Ind. Code §§ 35-33-7-1 to -2 (2017) (affording only Gerstein-style review); Me. R. Crim. P. 4A (requiring judicial review only of a sworn affidavit and only “when a defendant arrested without a warrant for any crime is not released from custody within 48 hours,” but also permitting the defendant to submit oral evidence undermining probable cause). and one (Delaware) goes to the far extreme of granting prosecutors full and unchecked control over the preliminary-hearing process via an information-bypass rule. 130 See infra note 290; Appendix at Table C (p. 1404). Finally, at the far other end of the spectrum, one state has flirted with but ultimately rejected a discretionary model in which judges can hold hearings even if an indictment or information has been filed, 131 Maryland has a statute expressly providing for a discretionary approach. See Md. Code Ann., Crim. Proc. § 4-103(2) (LexisNexis 2008) (“If the defendant is charged by grand jury indictment, the right of a defendant to a preliminary hearing is not absolute but the court may allow the defendant to have a preliminary hearing.”). But a contrary court rule creates an indictment-bypass regime, and the state supreme court has held that the rule takes precedence. See Md. R. 4-221(c)(1) (“A preliminary hearing may not be held if before the hearing . . . [a]n indictment is filed . . . .”); Marshall v. State, 420 A.2d 1266, 1269 (Md. Ct. Spec. App. 1980) (“[T]here . . . exists a conflict between the statute and the rule. As the rule was adopted subsequent to the enactment of the statutory provision, the rule prevails.” (citations omitted)), rev’d on other grounds, 434 A.2d 555 (Md. 1981). For discussion of the curious-but-common supremacy of court-made rules over legislative statutes, see infra section V.B and note 241. and an additional six have gone so far as to guarantee defendants a right to a pre­liminary hearing as a matter of course. 132 See Neb. Rev. Stat. § 29-1607 (2016); Okla. Stat. tit. 22, § 524 (West 2003); Wis. Stat. §§ 970.02, 971.02 (2018); N.D. R. Crim. P. 5(c); Pa. R. Crim. P. 540; Tenn. R. Crim. P. 5(e)(1), (4); see also infra Appendix at Table C. Three of these states layer papers review on top of that guarantee, thus adopting a belt-and-suspenders approach that arises from procedural peculiarities unique to each jurisdiction: In Pennsylvania, a defendant has a right to a preliminary hearing whether prosecuted by information or by indictment, except that a prosecutor can get permission from the state’s Supreme Court to indict a defendant without a preliminary hearing “in cases in which witness intimidation has occurred, is occurring, or is likely to occur.” Pa. R. Crim. P. 556.2 cmt.; see also Pa. R. Crim. P. 556. If, however, such a process is used, the defendant can then challenge any resulting indictment on the papers. See Pa. R. Crim. P. 556.4(B)(1)(b). In Nebraska, an indictment technically preempts a preliminary hearing. But a grand jury can only be summoned into being by a petition signed “by not less than ten percent of the registered voters of the county who cast votes for the office of Governor in such county at the most recent general election,” Neb. Rev. Stat. § 29-1401(3) (2016), a process so rare and cumbersome that preliminary hearings are effectively guaranteed. Cf. id. § 29-1607 (providing that “[n]o information shall be filed against any person for any offense until such person shall have had a preliminary examination”). In the rare event, however, that a grand jury is summoned and issues an indictment, that indictment is subject to dismissal if a court concludes that “the grand jury finding of probable cause is not supported by the record.” Id. § 29-1418(3). Finally, in Oklahoma, the defendant has a right to a preliminary hearing, Okla. Stat. tit. 22, § 524 (West 2003), and is further entitled to move to dismiss his case “for insufficient evidence . . . after preliminary hearing” if he can “establish beyond the face of the indictment or information that there is insufficient evidence to prove any one of the necessary elements of the offense for which [he] is charged,” State v. Delso, 298 P.3d 1192, 1193–94 (Okla. Crim. App. 2013) (emphasis added) (citing Okla. Stat. tit. 22, § 504.1(A)).
Note that Virginia guarantees preliminary hearings by statute, see Va. Code Ann. § 19.2-218 (2015) (“No person who is arrested on a charge of felony shall be denied a preliminary hearing . . . .”), but has narrowed that guarantee substantially, interpreting the statute to permit indictment bypass if the defendant was “indicted for charges distinct from those on which he was arrested but which arose out of the same course of events.” Seibert v. Commonwealth, 467 S.E.2d 838, 840 (Va. Ct. App. 1996). Virginia is thus not included among the six hearing-as-of-right states.

As this survey of state-by-state approaches makes clear, the structures of pretrial evidentiary review are exceptionally diverse, with preliminary hearings guaranteed in some states, nonexistent in others, and subject to varying modes of prosecutorial bypass in the rest. Note, however, that even when such a hearing takes place, the regulatory constraint it imposes on prosecutorial power depends on a final tier of the governing procedural framework, which determines how robustly the hearing will probe the allegations underlying the prosecution’s case. Specifically, procedural law determines the robustness of a given preliminary hearing in at least three important ways.

1. Standard of Review. — First, procedural law defines the standard of review employed at the hearing and thus sets the evidentiary threshold that the prosecutor must satisfy. The higher the standard, the more demanding the review—at least in theory. 133 Cf. Ortman, supra note 1, at 514 (“Strict charging standards are attractive to criminal justice policymakers for whom . . . prosecutorial constraint [is] important.”). But cf. C.M.A. McCauliff, Burdens of Proof: Degrees of Belief, Quanta of Evidence, or Constitutional Guarantees?, 35 Vand. L. Rev. 1293, 1327 & tbl.3 (1982) (reporting that judges, when surveyed, pegged probable cause at between a 10% and a 90% likelihood that a given allegation is true, with most answers clustered between 30% and 60%); Christopher Slobogin, Let’s Not Bury Terry: A Call for Rejuvenation of the Proportionality Principle, 72 St. John’s L. Rev. 1053, 1082 (1998) (“[P]robable cause . . . is the standard with which we are most familiar—except that we don’t really know what it means.”). States vary, however, in the standards they write into their subconstitutional procedural law: The majority borrow the probable cause standard from the Fourth Amendment, but others demand a “greater assurance of guilt” by requiring “evidence at the preliminary hearing that would warrant a reasonable jury in finding each of the elements of the offense” at a subsequent trial. 134 Stewart v. Abraham, 275 F.3d 220, 229–30 (3d Cir. 2001) (examining Pennsylvania state law and concluding that the “‘prima facie case’ standard” employed at preliminary hearings in that state is “intended to require different and greater assurance of guilt” than the more typical “‘probable cause’ standard”); see also 4 LaFave et al., supra note 73, § 14.3(a), at 356 (observing that the “dominant formulation of th[e] standard” applied at the hearing “directs the magistrate to determine whether ‘there is probable cause to believe that an offense has been committed and that the defendant committed it’” (quoting Alaska R. Crim. P. 5.1 and other states’ identically worded provisions)). With respect to papers review, at least one state goes so far as to require evidence that, if “taken in the light most favorable to the State,” would “tend to show [the] defendant committed the offense, beyond a reasonable doubt.” State v. Millette, 795 A.2d 1182, 1183 (Vt. 2002) (emphasis added).

2. Nature of Evidence. — Second, there is the question of how the prose­cutor will go about satisfying her burden, and more specifically, which witnesses she will be required to put forward. From the defendant’s perspective, the hearing will be most robust if the prosecutor must produce the principal witnesses upon whom she intends to rely at trial. In many states, however, the prosecutor can satisfy her burden of proof at the hearing entirely through hearsay testimony. 135 See Cal. Penal Code § 872(b) (2018); Del. Super. Ct. R. Crim. P. 5.1(a); Ga. Code Ann. § 24-1-2(d)(1) (2013); Ill. R. Evid. 1101(b)(3); Ky. R. Crim. P. 3.14(2); Md. R. 5-101(b)(7); Mont. R. Evid. 101(c)(3); Neb. Rev. Stat. § 27-1101(4)(b) (2016); N.H. R. Crim. P. 6(a)(4); N.J. R. Evid. 101(a)(2)(D) & cmt.; N.D. R. Crim. P. 5.1; 12 R.I. Gen. Laws § 12-12-1.9 (2002); S.C. R. Evid. 1101(d)(3); Wash. R. Evid. 1101(c)(3); Wis. Stat. § 970.038 (2018); Wyo. R. Evid. 1101(b)(3); see also Fed. R. Evid. 1101(d)(3); cf. 4 LaFave et al., supra note 73, § 14.4(b), at 383 (suggesting that “perhaps a majority” of states “start from the premise that the rules of evidence do not apply to the preliminary hearing”). In such a regime, a single police officer can simply take the stand and summarize the most inculpatory portions of the case file, thus shielding potentially weak witnesses from cross-examination and perhaps sanitizing their accounts in the process. Under this approach, our hypothetical armed robbery defendant’s attorney, for example, might find herself cross-examining an officer who testifies simply that “a gentleman saw the defendant commit the robbery” and who conspicuously omits the fact that the gentleman in question is blind, a convicted felon, and biased against the defendant. 136 Note, moreover, that due to the lack of pretrial discovery and insufficient investigative budgets in many jurisdictions, the defense attorney may have no idea about the witness’s deficiencies if the witness does not testify at the preliminary hearing. See Donald J. Farole, Jr. & Lynn Langton, U.S. Dep’t of Justice, Bureau of Justice Statistics, County-Based and Local Public Defender Offices, 2007, at 1 (2010), [] (reporting that “40% of all county-based public defender offices had no investigators on staff”); Grunwald, supra note 121, at 779–80 (describing discovery rules). Other states, however, take steps to restrict such obfuscation: Some give judges discretion to accept or reject hearsay at the hearing; 137 See, e.g., Iowa Ct. R. 2.2(4)(b) (permitting hearsay if the judge finds there to be “a substantial basis for believing the source of the hearsay [is] credible and for believing that there is a factual basis for the information furnished”); La. Code Evid. Ann. art. 1101(B)(4) (2017) (“[T]he court may consider evidence that would otherwise be barred by the hearsay rule.”); Or. Rev. Stat. § 135.173 (2015) (prohibiting hearsay unless the judge “determines that it would impose an unreasonable hardship on one of the parties or on a witness to require that the primary source of the evidence be produced at the hearing”). others permit hearsay only in narrowly limited circumstances; 138 See, e.g., Idaho R. Crim. P. 5.1(b) (“Hearsay . . . may be admitted to show the following: (1) the existence or nonexistence of business or medical facts and records, (2) judgments and convictions of courts, (3) ownership of real or personal property, and (4) reports of scientific examinations of evidence by state or federal agencies . . . .”); Kan. Stat. Ann. § 22-2902(3) (2007) (“When the victim . . . is a child less than 13 years of age, the finding of probable cause . . . may be based upon hearsay evidence in whole or in part presented at the preliminary examination by means of statements made by [the] child . . . on a videotape recording or by other means.”); Nev. Rev. Stat. § 171.196(6) (2015) (permitting “[h]earsay evidence . . . at a preliminary exami­nation . . . only if the defendant is charged with one or more of [certain enumerated] offenses”). and others ban it altogether. 139 See, e.g., Ohio Rev. Code Ann. § 2937.11(A)(3) (West 2006) (“The hearing shall be conducted under the rules of evidence prevailing in criminal trials generally.”); S.D. Codified Laws § 23A-4-6 (2018) (stating that “[t]he rules of evidence shall apply” at the preliminary hearing); Tex. Code Crim. Proc. art. 16.07 (2017) (“The same rules of evidence shall apply to and govern a trial before an examining court that apply to and govern a final trial.”); Va. Code Ann. § 19.2-183(B) (2015) (“At the hearing the judge shall . . . hear testimony . . . in accordance with the rules of evidence applicable to criminal trials . . . .”); Myers v. Commonwealth, 298 N.E.2d 819, 824 n.6 (Mass. 1973) (“[T]he rules of evidence at the preliminary hearing should in general be the same rules that are applicable at the criminal trial.”); cf. Alaska R. Crim. P. 5.1(d) (“At the preliminary examination, the admissibility of evidence other than written reports of experts shall be governed by the Alaska Rules of Evidence.”). Indeed, some states go so far as to give the defendant himself a right to call the prosecutor’s witnesses to the stand in order to question them directly. 140 See, e.g., Ex parte Lankford, 20 So. 3d 843, 844– 46 (Ala. Crim. App. 2009) (per curiam) (holding that the defendant has a right to subpoena “the alleged victim” to testify at a preliminary hearing); Wyrick v. Dist. Court of Mayes Cty., 839 P.2d 1376, 1377 (Okla. Crim. App. 1992). Notably, jurisdictions also vary on the extent to which defendants are permitted to present any evidence at a preliminary hearing at all. Compare, e.g., Haw. R. Penal P. 5(c)(4) (“The prosecution and the defendant may introduce evidence and produce witnesses, who shall be subject to cross-examination.”), with Del. Ct. Com. Pleas Crim. R. 5.1 (“The defendant . . . may, subject to reasonable limitations imposed by the Court, introduce evidence in defendant’s behalf.”), N.Y. Crim. Proc. Law § 180.60(7) (McKinney 2007) (“Upon request of the defendant, the court may, as a matter of discretion, permit him to call and examine other witnesses or to produce other evidence in his behalf.”), and Md. R. 4-221(d) (“The defendant is entitled to cross-examine witnesses but not to present evidence.”).

3. Consequences of Overreaching. — Finally, and perhaps most importantly, procedural law determines the consequences that follow if the judge finds at the end of the hearing that the prosecutor’s charges fail to pass muster. In many jurisdictions those consequences are almost nonexistent, as the prosecutor is free simply to refile the rejected charges, perhaps before a more amenable judge—an approach that leaves her free to overreach again and again, until her charges stick or she gives up. 141 See Ala. R. Crim. P. 5.4(d); Alaska R. Crim. P. 5.1(h); Ariz. R. Crim. P. 5.4(b)–(d) & cmt.; Del. Super. Ct. R. Crim. P. 5.1(b); Fla. R. Crim. P. 3.133(b)(5); Iowa Ct. R. 2.2(4)(e); La. Code Crim. Proc. Ann. art. 386 (2017); Md. R. 4-221(e); Miss. R. Crim. P. 6.2(g); Mont. Code Ann. § 46-10-203(2) (2017); N.H. R. Crim. P. 6(a)(6); N.M. R. 6-202(D)(1); N.C. Gen. Stat. § 15A-612(b) (2017); Ohio R. Crim. P. 5(B)(5); Okla. Stat. Ann. tit. 22, § 501 (West 2003); S.C. R. Crim. P. 2(c); S.D. Codified Laws § 23A-4-7; Tenn. R. Crim. P. 5.1(c); Utah R. Crim. P. 7(i)(3); Vt. R. Crim. P. 5(c); W. Va. R. Crim. P. 5.1(b); Wyo. R. Crim. P. 5.1(c); State v. Rubek, 371 N.W.2d 115, 117 (Neb. 1985); State v. Farrad, 753 A.2d 648, 660 (N.J. 2000); Herrington v. Commonwealth, 781 S.E.2d 561, 563 (Va. 2016); see also Fed. R. Crim. P. 5.1(f). Some states do not permit dismissal even upon an adverse judicial finding, requiring only that the defendant be released from pretrial custody (though not from other restraints on his liberty). See, e.g., N.J. Ct. R. 3:4 -3. As for forum shopping, note that states may (but often do not) prohibit the tactic by requiring that any “subsequent preliminary examination must be held before the same judicial officer.” Mich. R. Crim. P. 6.110(F). By contrast, some states require a prosecutor to seek the court’s permission before refiling charges that have been rejected, 142 See, e.g., Nev. Rev. Stat. § 173.035 (2015) (“If . . . upon the preliminary examination the accused has been discharged . . . the district attorney may, upon affidavit of any person who has knowledge of the commission of [the] offense, and . . . by leave of the court first had, file an[other] information . . . .”); Wash. R. Cts. Limited Jurisdiction 3.2.1(g)(5) (“If a preliminary hearing on the felony complaint is held and the court finds that probable cause does not exist, the charge shall be dismissed, and may be refiled only if a motion to set aside the finding is granted by the superior court.”). while others go so far as to ban such refiling altogether unless the prosecutor can show that she has obtained new evidence that she either did not or could not have obtained the first time around. 143 See, e.g., Cal. Penal Code §§ 871, 1387(a)(1) (2018) (permitting refiling if “substantial new evidence has been discovered by the prosecution which would not have been known through the exercise of due diligence at, or prior to” the preliminary hearing dismissal); Mich. R. Crim. P. 6.110(F) (requiring “the prosecutor [to] present additional evidence to support the charge” at a subsequent preliminary hearing on refiled charges); 12 R.I. Gen. Laws § 12-12-1.10 (2002) (barring refiling absent “a finding [by the court] of mistake, inadvertence, surprise, excusable neglect, the discovery of new evidence which by due diligence could not have been discovered at the time the hearing on probable cause was held, or any other reason justifying the relief”); Wis. Stat. § 970.04 (2018) (“If a preliminary examination has been had and the defendant has been discharged, the district attorney may file another complaint if the district attorney has or discovers additional evidence.”).

In sum, if one were to focus solely on the lack of constitutionally grounded regulation in this arena it would be easy to conclude that the job of assessing the facts underlying an alleged criminal offense falls, to quote the Supreme Court, within “a ‘special province’ of the Executive,” such that the prosecutor’s charging decisions must rest inevitably and “‘entirely in [her] discretion.’” 144 United States v. Armstrong, 517 U.S. 456, 464 (1996) (first quoting Heckler v. Chaney, 470 U.S. 821, 832 (1985); then quoting Bordenkircher v. Hayes, 434 U.S. 357, 364 (1978)). In truth, however, a prosecutor’s ability to inflate the factual allegations against the defendant—and to thereby gener­ate outsized plea bargaining leverage—depends on the subconsti­tutional procedural law of the state in which the prosecution occurs. Some states provide no pretrial evidentiary review at all, others guarantee something approxi­mating a minitrial, and a full assortment of alternatives fall in between.

B. Legal Overreaching: The Law of Summary Dismissal and Bills of Particulars

Unlike factual overreach, legal overreach arises relatively infrequently in criminal prosecutions, for a simple reason: In most jurisdictions, the well-settled substantive criminal law defines liability so expansively that it covers most alleged criminal misconduct many times over, relieving prosecutors of any need to stretch for legally applicable charges. 145 See Robert H. Jackson, The Federal Prosecutor, 24 J. Am. Judicature Soc’y 18, 19 (1940) (“With the law books filled with a great assortment of crimes, a prosecutor stands a fair chance of finding at least a technical violation of some act on the part of almost anyone.”); cf. James M. Burnham, Why Don’t Courts Dismiss Indictments? A Simple Suggestion for Making Federal Criminal Law a Little Less Lawless, 18 Green Bag 2d 347, 358 (2015) (“[T]here is not much gray about what constitutes bank robbery or drug possession.”). Still, legal overreaching is not a phantom menace, 146 Cf. Burnham, supra note 145, at 358 & n.30 (citing examples illustrating that “for many areas of federal criminal law” in particular, “the government has a long track record of pursuing aggressive, questionable legal theories”). and when it occurs it can offer prosecutors substantial leverage. Consider, for example, our hypo­thetical armed robbery defendant, who when we first encountered him was facing not only a pile of robbery-like charges (theft, assault, brandishing a firearm) but also a far more serious charge of kidnapping—premised solely on the allegation that, in the course of the robbery, he told the victim to move a few steps to the left, out from under a streetlamp. 147 See supra text accompanying note 36. Putting aside for the moment the factual question of whether the defendant ever issued such a command (or committed the robbery), the prosecutor’s ability to threaten him with a hefty kidnapping charge depends on the answer to a straightforward question of law: Does this brief alleged restraint of the victim’s freedom of movement constitute a kidnapping?

That legal question could well be unresolved in the particular jurisdiction at hand. 148 Cf. Melanie A. Prince, Comment, Two Crimes for the Price of One: The Problem with Kidnapping Statutes in Tennessee and Beyond, 76 Tenn. L. Rev. 789, 789–90 (2009) (observing that “statutes and case law in several jurisdictions” across the country “could allow prosecution of criminal defendants for both kidnapping and the underlying offense” if the conduct involved “any asportation, however slight” or “even simple” or temporary detention); id. at 789 (“As a result, criminal defendants could face two convictions for essentially one action, and the severe sentences that were originally intended to punish ransom-kidnappers would be applied to those guilty of a lesser crime.”). And if it is, uncertainty over how courts will resolve the issue creates a now-familiar opportunity for charge bargaining and manipulation: To eliminate the risk of a lengthy kidnapping sentence, the defendant might jump at the chance to plead guilty to a lesser offense—including, perhaps, to armed robbery, the charge the prosecutor likely preferred all along. 149 Cf. Gold et al., supra note 20, at 1642 (“Criminal laws often have unclear boundaries. This lack of clarity allows prosecutors to push more aggressive legal theories.”). For other discussions of uncertainty yielding prosecutorial leverage, see supra notes 92–94 and accompanying text (discussing discretionary sentencing); supra notes 98–104 and accompanying text (discussing the inherent uncertainty of jury verdicts). To the extent that undue leverage arising from such uncertainty causes concern, a potential regulatory solution should by now be familiar as well: Resolve the uncertainty up front, by permitting the judge to rule on the legal issue before plea negotiations are over—that is to say, at some point before trial. 150 Cf. supra text accompanying note 94 (discussing structured sentencing discretion as a mode of clarifying uncertainty); supra section III.A (discussing pretrial evidentiary review in similar terms).

Here again, the constitutional law of criminal procedure falls short, insofar as it guarantees defendants judicial review of the prosecutor’s legal theory of the case only if the defendant takes the case to trial, by which point any hope of checking legal overreach in the plea negotiation process is lost. 151 See Jackson v. Virginia, 443 U.S. 307, 314, 319–20 (1979) (guaranteeing judicial review to ensure “that every element of the crime has been established”). While Jackson requires judges to review the evidence adduced at trial and thus creates a limited check on factual overreaching, the deferential nature of that factual inquiry renders the precedent perhaps more significant as a check on legal overreaching, as it logically requires judges to resolve any ambiguity in a charged crime’s legal definition before resolving whether that definition is satisfied by the facts at hand. See, e.g., Carrington v. Commonwealth, 721 S.E.2d 815, 817–18 (Va. Ct. App. 2012) (interpreting boundaries of statutory elements in the context of Jackson review). But where constitutional law falls short, subconstitutional law fills the gap, in this instance through a little-studied procedural device known as a motion for summary dismissal. 152 See James Fallows Tierney, Comment, Summary Dismissals, 77 U. Chi. L. Rev. 1841, 1841–42, 1849–50 (2010) (describing “the near-total scholarly silence” concerning summary dismissals). As James Tierney explains in a comment offering the most extensive analysis of this device, a motion for summary dismissal in a criminal case mirrors the civil system’s more familiar motion to dismiss for failure to state a claim. 153 Tierney, supra note 152. Tierney’s thorough analysis focuses exclusively on the Federal Rules of Criminal Procedure, as does the article that was, at the time Tierney wrote, “the only other scholarly treatment of this issue.” Id. at 1841–42 & n.9 (citing James M. Shellow & Susan W. Brenner, Speaking Motions: Recognition of Summary Judgment in Federal Criminal Procedure, 107 F.R.D. 139 (1985)). For more recent treatments, also focusing solely on the federal rules, see Burnham, supra note 145, at 348–49; Gold et al., supra note 20, at 1641–42. As such, the device forces the government to “prove its case as a matter of law” at the outset of the proceedings and thereby allows the defendant to “expedite resolution of legal questions that would otherwise be litigated” at trial. 154 Tierney, supra note 152, at 1841, 1849. In other words, summary dismissals move the judicial review guaranteed by the Constitution up to a point in the proceedings when such review could actually be useful. 155 Cf. Gold et al., supra note 20, at 1642 (arguing that increasing the usage of motions for summary dismissal “would provide an opportunity for defendants to test the validity of the prosecutors’ legal theories”).

Notably, in contrast to the many other thickets of subconstitutional law examined in this Article, the procedural landscape here is mercifully straightforward, as motions for summary dismissal are “specifically recog­nized in the statutes or court rules of every state” in the country—that is to say, in each state’s subconstitutional procedural law. 156 5 LaFave et al., supra note 73, § 19.3(a), at 286. But cf. 4 id. § 15.5(a), at 631–35 (describing the somewhat-more-varied summary-dismissal rules in the federal system); Tierney, supra note 152, at 1845, 1853–54 (noting that while most federal courts agree that “a district judge can hear the motion, resolve the legal question, and dismiss if the government has no case to prove,” the Eighth Circuit does not allow for summary dismissals (citing United States v. Nabors, 45 F.3d 238 (8th Cir. 1995))). That consistency, however, tells only half the story, for the difficulty defendants typically face when pushing back against legal overreach is not finding a procedural device through which to mount their challenge but rather pinning the prosecutor to a specific legal theory of liability in the first place. Criminal charging instruments, after all, are much sparser than civil complaints, often alleging little more than the time and place of the offense. 157 Cf. Gold et al., supra note 20, at 1640–41 (contrasting criminal “[c]harging instruments [that] generally do not need to allege specific facts” to the heightened pleading standards in federal civil cases (citing Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009))); infra Appendix at Table D (discussing charging standards). And while pretrial evidentiary hearings can help determine whether the prosecutor has any factual basis to file a charge, they rarely examine every potential charge the eventual trial could entail, nor do they typically bind the prosecutor to the factual narrative adduced at the hearing. 158 Cf. 4 LaFave et al., supra note 73, § 14.1(c), at 317 (noting “that there [can] be some inconsistency between [witnesses’] trial testimony and their previous statements [at a preliminary hearing]”); Frank W. Miller, Prosecution: The Decision to Charge a Suspect with a Crime 67–68 (Frank J. Remington ed., 1970) (observing that prosecutors “normally follow a policy of revealing no more evidence than is necessary to obtain a bindover” of the case from the preliminary hearing to the formal charging stage). The precise factual predicate for the charges, in other words, is often hazy, which gives prosecutors leeway to obfuscate pretrial litigation of the criminal law’s boundaries. Our hypothetical prosecutor, for example, may not be willing to stipulate that her kidnapping charge rests exclusively on the defendant’s alleged “streetlamp command,” either because she doesn’t know for sure what her witnesses will say at trial and wants to keep her options open or because she recognizes that her legal theory is novel and wants to insulate it and the leverage it affords from pretrial adjudication.

In other words, effective constraints on legal overreach often require some regulatory mechanism to force prosecutors to specify their legal theory early on—which subconstitutional procedural law affords by way of a motion for a bill of particulars. If granted, such a motion compels the prosecutor to answer questions posed by the defense “concerning the events cited in the charge” and treats those answers as binding with respect to “the government’s case at trial.” 159 5 LaFave et al., supra note 73, § 19.4(a), at 358–59. To see the impact of such a device in practice, note the dilemma our hypothetical prosecutor faces if she must identify with specificity the actus reus of her kidnapping charge: She can either acknowledge that the defendant’s streetlamp command is indeed the crux of the charge (and face the ensuing legal challenge head on), or she can commit herself to a factual predicate that sits more comfortably within the kidnapping statute (and run the risk that the evidence she can actually produce at trial may not support that newly refined allegation).

This choice, however, will be forced only if the governing procedural law requires the prosecutor to issue a bill of particulars—and here again, states vary in their approaches.

Figure 15: Variability Across States in the Law of Bills of Particulars 160 For sources, see infra Appendix at Table D.


As the chart above shows, most states grant judges discretion to call for a bill of particulars, with the exercises of such discretion potentially ex­tend­ing “from one extreme to the other.” 161 United States v. Callahan, 18 F.R.D. 486, 489 (W.D. Wash. 1955); see also 5 LaFave et al., supra note 73, § 19.4(b), at 367. Of course, regulatory constraints grounded in discretion will turn on how judges use that discretion. 162 Cf. supra text accompanying notes 50–51 (discussing equitable-failsafe severance); supra text accompanying notes 90–92 (discussing discretionary sentencing); supra text accompanying note 120 (discussing discretionary preliminary hearings). But as the chart above reflects, states can also adopt more categorical approaches, which some do—in both directions. Specifically, ten states have eliminated the bill of particulars altogether, leaving prosecutors free either to stipulate to facts that support pretrial adjudication of legal issues or to evade such review until trial. 163 See infra Appendix at Table D. Prosecutors may sometimes want to facilitate a pretrial legal ruling on their theory of a case in order to confirm its validity prior to committing resources to a trial, a route they may be particularly likely to pursue if the case at hand is intended as a test vehicle for a novel legal theory. Cf. Crespo, Regaining Perspective, supra note 13, at 2032–33 (discussing prosecutors’ incentives to pursue favorable appellate precedent). But two other states (New York and Ohio) give defendants the right to demand a bill of particulars that “specifically [states] the nature of the offense charge[d]” and “the conduct of the defendant alleged to constitute the offense.” 164 Ohio R. Crim. P. 7(E) (requiring the prosecuting attorney to “furnish the defendant” with such information “[w]hen the defendant makes a written request”); see also N.Y. Crim. Proc. Law § 200.95 (McKinney 2007) (stating that “[u]pon a timely request for a bill of particulars by a defendant against whom an indictment is pending, the prosecutor shall” file one stating “the substance of [the] defendant’s conduct encompassed by the charge which the people intend to prove at trial on their direct case”). In these states, as the discussion above makes clear, the subconstitutional procedural law of summary dismissal and bills of particulars creates a genuine mechanism for checking legal overreach.

C. Equitable Overreaching: The Law of Equitable Dismissal

For critics of American plea bargaining, the problem of equitable overreach captures perhaps the core concern with plea bargaining more generally: Defendants are sometimes threatened with charges that are simply more serious than their alleged conduct seems to deserve. 165 See supra notes 28, 37 and accompanying text. Take our hypothetical defendant: Even if one accepts that his alleged conduct technically meets the legal elements of a kidnapping, one might appropriately ask whether threatening him with such a charge—and with the decades of imprisonment it could entail—is fair, just, and equitable under the circumstances. 166 Indeed, one might ask the same question with respect to the sentences typically meted out for armed robbery. Cf. Pfaff, Locked In, supra note 17, at 185–202 (stressing the need to reevaluate societal approaches to punishing violent crime). As Professor Josh Bowers explains, a prosecutor tasked with deciding what charges to bring and what plea deals to offer cannot escape the exercise of equitable judgment, which is both “necessary and desirable” in a world where criminal codes are too expansive and too severe to permit prosecuting every alleged wrongdoer to the hilt. 167 Josh Bowers, Legal Guilt, Normative Innocence, and the Equitable Decision Not to Prosecute, 110 Colum. L. Rev. 1655, 1664 (2010) [hereinafter Bowers, Normative Innocence]; see also Davis, supra note 4, at 25 (“All governments in history have been governments of laws and of men. Rules alone, untempered by discretion, cannot cope with the complexities of modern government and of modern justice.”); Langer, supra note 20, at 242 (“In [the current] political and institutional context, . . . prosecutors may not merely rely on the fact that the legislature has passed a crime to justify their charging decisions. They also have to make a normative decision on whether the conduct in question is worth prosecuting . . . .”). The concern, in other words, is not that prosecutors’ equitable discretion exists, but rather that it so often seems to be exercised without regard to identified or identifiable standards, and that it operates seemingly free of external constraint—with observably problematic results. 168 See supra notes 15–17, 26–30 and accompanying text.

As Bowers notes, an obvious corrective to this problem would be to “apportion equitable decisionmaking power” between prosecutors and some other actor, an intervention that could be accomplished easily enough by allowing judges to review the basic fairness of prosecutors’ charging decisions. 169 Bowers, Normative Innocence, supra note 167, at 1680. Professor Bowers proposes a different institutional response, urging the creation of “a lay body—analogous to the grand jury—that would presume evidentiary sufficiency and would, instead, address only normative questions of whether charges equitably ought to be filed.” Id. at 1725; see also Josh Bowers, The Normative Case for Normative Grand Juries, 47 Wake Forest L. Rev. 319, 320–21 (2012). But see supra note 105 (citing and discussing extensive literature on the shortcomings of grand jury review). And yet, while judicial oversight is an accepted staple of checks and balances in other settings, the suggestion that judges might review prosecutors’ equitable discretion will strike students of the criminal justice system as curious—perhaps even blasphemous. After all, the constitutional law of criminal procedure treats “the power to prosecute” as “one of the core powers of the Executive Branch,” and accordingly sees “‘the decision whether or not to prosecute, and what charge to file or bring’” as lying “‘entirely in [prosecutors’] discretion.’” 170 United States v. Armstrong, 517 U.S. 456, 464, 467 (1996) (quoting Bordenkircher v. Hayes, 434 U.S. 357, 364 (1978)); see also Inmates of Attica Corr. Facility v. Rockefeller, 477 F.2d 375, 379–80 (2d Cir. 1973); United States v. Cox, 342 F.2d 167, 171 (5th Cir. 1965) (“It follows, as an incident of the constitutional separation of powers, that the courts are not to interfere with the free exercise of the discretionary powers of the attorneys of the United States in their control over criminal prosecutions.”); cf. Bowers, Normative Innocence, supra note 167, at 1686 (“In the charging context, courts and scholars take it as something of an article of faith that the prosecutor should enjoy principal—or even exclusive—authority.”); id. at 1659 (asserting that “prosecutors enjoy almost unbridled equitable discretion”); Meares, supra note 17, at 862 (“So long as the prosecutor has probable cause to believe that the accused committed an offense, the prosecutor is entitled to bring the charge [and will] rarely [be] second-guessed by the courts.”).

But while this division of institutional responsibilities is dictated by Supreme Court doctrine, it is not a command of natural law—nor does it bind the states, which are free to assign their courts and their prosecutors shared responsibility when it comes to ensuring that criminal charges are not only factually and legally viable, but also fair and just. 171 Cf. Dreyer v. Illinois, 187 U.S. 71, 84 (1902) (“Whether the legislative, executive and judicial powers of a State shall be kept altogether distinct and separate, or whether persons . . . belonging to one department may, in respect to some matters, exert powers which, strictly speaking, pertain to another department of government, is for the determination of the State.”). And indeed, one-third of the states have done just that, enacting through their subconstitutional law a set of procedural mechanisms that authorize courts to dismiss charges prior to trial purely on equitable grounds. Only recently examined in the scholarly literature, these procedural devices exist in one of two forms. 172 See Anna Roberts, Dismissals as Justice, 69 Ala. L. Rev. 327, 330 (2017) (offering the first detailed scholarly treatment of a topic previously suffering from longstanding “neglect on the part of scholars”). The first, adopted in Hawaii, Maine, New Jersey, and Pennsylvania, tracks a proposal first laid out in the Model Penal Code, which contains a “De Minimis Infractions” provision under which judges can “dismiss a prosecution” before trial if they find “that the defendant’s conduct” was “too trivial to warrant the condemnation of conviction.” 173 Model Penal Code § 2.12 (Am. Law Inst. 1962); see also Haw. Rev. Stat. Ann. § 702-236 (LexisNexis 2016); Me. Rev. Stat. Ann. tit. 17-A, § 12 (2006); N.J. Stat. Ann. § 2C:2-11 (West 2015); 18 Pa. Stat. and Cons. Stat. Ann. § 312 (West 2015). The Model Penal Code also asks courts to consider whether the alleged conduct “was within a customary license or tolerance” or “presents such other extenuations that it cannot reasonably be regarded as envisaged by the legislature in forbidding the offense,” and makes dismissal mandatory if any of the three statutory prongs are satisfied. Model Penal Code § 2.12. With the exception of Pennsylvania, however, all of the adopting states have converted the provision to a discretionary one. See Stanislaw Pomorski, On Multiculturalism, Concepts of Crime, and the “De Minimis” Defense, 1997 BYU L. Rev. 51, 82–83. Moreover, some of the states adopting the provision have further narrowed it through judicial construction to exclude certain enumerated categories of offenses (for example, felonies). See Roberts, supra note 172, at 356–60. Meanwhile, the second approach—adopted in fifteen states—seemingly goes even further, empowering judges to dismiss charges whenever they conclude “that such dismissal will serve the ends of justice.” 174 E.g., Idaho Crim. R. 48(a)(2) (“The court . . . may dismiss a criminal action . . . if the court concludes that dismissal will serve the ends of justice and the effective administration of the court’s business.”); see also Roberts, supra note 172, at 332–33 & n.21 (citing Alaska R. Crim. P. 43(c); Ariz. R. Crim. P. 16.6(b) (abrogated Aug. 31, 2017); Cal. Penal Code § 1385(a) (2018); Conn. Gen. Stat. § 54-56 (West 2009); Idaho Code § 19-3504 (2017); Idaho Crim. R. 48(a)(2); Iowa Ct. R. 2.33(1); Minn. Stat. Ann. § 631.21 (West 2009); Mont. Code. Ann. § 46-13-401(1) (West 2015); N.Y. Crim. Proc. Law §§ 170.40, 210.40 (McKinney 2007); Ohio R. Crim. P. 48(B); Okla. Stat. Ann. tit. 22, § 815 (West 2003); Or. Rev. Stat. § 135.755 (2003); Utah R. Crim. P. 25(a); Vt. R. Crim. P. 48(b)(2); Wash. R. Crim. P. 8.3). In some states, this authority has been structured by appellate decisions that prescribe certain factors to guide trial courts’ discretion. See State v. Sauve, 666 A.2d 1164, 1167–68 (Vt. 1995) (surveying jurisdictions and listing factors employed in three other states). But in others, the authority exists as an essentially open-ended warrant to keep prosecutors’ equitable overreach in check. Cf. 4 LaFave et al., supra note 73, § 13.2(c), at 151–52 (“While some states have construed such a provision to mean that the court may dismiss a prosecution only upon a showing of arbitrary action or government misconduct, elsewhere the trial court’s authority has not been so limited . . . .” (footnotes omitted)).

Figure 16: Variability Across States in the Law of Equitable Dismissal 175 For sources, see supra notes 172–174.


As Professor Anna Roberts observes in her recent examination of this topic, these equitable dismissal provisions offer “an important set of tools” that judges can use to act as “a check on the executive.” 176 Roberts, supra note 172, at 330. Of course, how judges use the discretion these devices afford is an important part of the equation. See supra text accompanying notes 50, 94, 120; infra section V.B. Plea bargaining scholars have long advocated precisely such an intervention, which one-third of the states already provide through their subconstitutional procedural law.

IV. Sliding Down

Up to this point we have examined how procedural law regulates prosecutors’ ability to maximize their charge-bargaining leverage through various modes of piling on and overreaching. In order for the prosecutor to capitalize on that leverage, however, she needs to be able to replace her inflated charges with some lower set of charges that can be offered to the defendant as an inducement to plead guilty: It is the differential between the threat and the offer that creates the leverage. 177 See Gazal-Ayal, supra note 34, at 2298–99 (observing that a prosecutor “can assure a conviction” in “almost every case, even a very weak one . . . by offering the defendant a substantial discount”); see also supra note 34. Indeed, if the prosecutor’s ability to slide between charges is restricted, she may be reluctant to pile on or overreach in the first place, out of fear that she will incur backlash from the jury. After all, jurors have unreviewable and plenary power to acquit, which means they can reject not just charges that they think are factually unsupported but also charges that they deem too severe—whether because they disagree with the governing substantive law or with the prosecutor’s decision about what charges to file. 178 See, e.g., Jeffrey Abramson, Two Ideals of Jury Deliberation, 1998 U. Chi. Legal F. 125, 145–50 (discussing alternative normative conceptions of jury nullification); see also Jack B. Weinstein, Considering Jury “Nullification”: When May and Should a Jury Reject the Law to Do Justice, 30 Am. Crim. L. Rev. 239, 240 (1993) (offering one federal trial judge’s view that “[n]ullification is but one legitimate result in an appropriate constitutional process,” such that “[w]hen juries refuse to convict on the basis of what they think are unjust laws, they are performing their duty as jurors”). Notably, the extent to which a jury is made aware of its unqualified power to acquit is a function of the subconstitutional procedural law relating to instructions on this issue, which once again varies across jurisdictions: Most tell juries only of their power to reject charges that are factually unsupported, but others alert juries to their broader power of acquittal. Compare State v. Ragland, 519 A.2d 1361, 1372 (N.J. 1986) (capturing the majority view that “[j]ury nullification is an unfortunate but unavoidable power [that] should not be advertised”), with Hebron v. State, 627 A.2d 1029, 1036 (Md. 1993) (holding that a state constitutional provision declaring that “the Jury shall be the Judges of Law” means that jurors have the authority to “resolv[e] conflicting interpretations of the law of the crime and [to] determin[e] whether that law should be applied in dubious factual situations”). A jury, in other words, can reject a prosecutor’s charges precisely because they are factually, legally, equitably, or numerically inflated, which means that a prosecutor who abuses such tactics runs the risk of an overplayed hand.

The prosecutor can mitigate or eliminate that risk, however, if she can easily replace her overplayed hand with a less audacious one. This is not a novel idea. On the contrary, plea bargaining scholars have long noted that restricting a prosecutor’s ability to reduce a defendant’s sentencing exposure can make it more “costly to choose unrealistically high charges” at the outset of the negotiations. 179 Wright & Miller, supra note 10, at 86; see also Oren Bar-Gill & Oren Gazal Ayal, Plea Bargains Only for the Guilty, 49 J.L. & Econ. 353, 357–58 (2006) (offering a formal model); Russell D. Covey, Fixed Justice: Reforming Plea Bargaining with Plea-Based Ceilings, 82 Tul. L. Rev. 1237, 1243–56 (2008) (noting that limits on sliding down “reduce prosecutorial incentives to overcharge criminal defendants by eliminating the bargaining leverage that can be obtained through strategic overcharging”); Gazal-Ayal, supra note 34, at 2313–16; James Vorenberg, Decent Restraint of Prosecutorial Power, 94 Harv. L. Rev. 1521, 1560–61 (1981) (suggesting that implementing a “prescribed sentencing concession of ten or twenty percent of the sentence received for a guilty plea” would “avoid much of the injustice of giving prosecutors heavy leverage to discourage even those who have a good defense from exercising their right to trial”). Most of these authors focus on how such restrictions could curb factual overreach, but as stated in the text, the same logic applies to piling on, legal overreach, and equitable overreach, insofar as a jury could react negatively to each tactic. In the absence of such restrictions, the prosecutor is free to file inflated charges without fear of overshooting the mark, because she knows that she will have the flexibility to slide down incrementally from her opening salvo to whatever lesser set of charges ultimately prompts the defendant to cave. 180 Cf. Gazal-Ayal, supra note 34, at 2307 (observing that when the ability to slide down is “unlimited,” the underlying “strength” of the initial charges “becomes less important to the prosecution,” which is incentivized to file even “weak” charges in order to generate leverage). But if the prosecutor’s initial charges somehow become “stickier,” she will be forced to screen away excessively inflated charges up front. Otherwise, she runs the risk that the defendant will call her bluff and thereby make her choose between abandoning those charges altogether or instead pressing ahead with a resource-intensive trial on charges the jury could well reject. 181 See Gazal-Ayal, supra note 34, at 2300; see also Covey, supra note 179, at 1245 (noting that restrictions on charge reductions “directly impact prosecutorial screening practices, creating strong incentives to dismiss weak [charges] rather than try them”); Wright & Miller, supra note 10, at 44 (“More careful selection of cases would make it possible to stick with the initial charges, even in front of a judge or jury.”). By de­ter­mining just how sticky those initial charges will be, the subconstitutional procedural law of amendment, dismissal, and lesser offenses regulates this final component of charge-bargaining power.

A. Charge Sliding: The Law of Amendment and Dismissal

Scholars who aim to curb charge bargaining through restrictions on sliding down tend to propose one of two interventions. The first, offered by Professors Ronald Wright and Marc Miller, calls for a self-imposed ban on charge bargaining: Head prosecutors, they say, should simply prohibit their subordinates from changing charges once they have been filed. 182 See Wright & Miller, supra note 10, at 46, 49 (advocating a “ban on charge reductions” under which “the prosecutor spends resources up front in assessing the case, and then lives with the consequences” of her initial charging decision “through final adjudication”). The merits of this proposal aside, however, it is hard to call it “regulation” of prosecutorial power, as the proposal depends entirely on the voluntary compliance of the regulated actors (the prosecutors). 183 See id. at 32 (proposing intervention dependent on “internal enforcement mecha­nisms”). On the dangers of relying exclusively on prosecutorial self-regulation, see, for example, Davis, supra note 4, at 211–12 (“The reasons for a judicial check of prosecutors’ discretion are stronger than for such a check of other administrative discretion that is now traditionally reviewable [by courts]. Important interests are at stake. Abuses are common. The questions involved are appropriate for judicial determination. And much injustice could be corrected.” (emphasis omitted)); see also Crespo, Systemic Facts, supra note 6, at 2059–65 (discussing the institutional advantages of courts with respect to systemic regulation of law enforcement actors). By contrast, the second proposal, advanced by Professors Oren Gazal-Ayal and Russell Covey, calls for judges to limit prosecutors’ leverage by limiting the differentials between the sentences that judges themselves impose in cases resolved by trial and in cases resolved by plea. 184 See Covey, supra note 179, at 1269 (proposing a regime that “guarantee[s] defendants that they will not receive a sentence following a trial conviction that is more severe than any plea offer made to them, adjusted upward by [an] appropriate fixed discount”); Gazal-Ayal, supra note 34, at 2313 (proposing that courts “reject [a] plea bargain whenever the suggested sentence is significantly lower than the [anticipated] post-trial sentence”). But while this approach admirably leans on judicial oversight to cabin prosecutorial power, it ultimately does little, as Gazal-Ayal acknowledges, to guard against the chief concern—charge bargaining—because prosecutors would still be able to manipulate the defendant’s sentencing exposure by manipulating the underlying offenses. 185 See Gazal-Ayal, supra note 34, at 2340 (acknowledging that “[c]harge bargains present a significant concern” for the proposed reform); see also Bar-Gill & Gazal Ayal, supra note 179, at 360 (observing that the screening effect associated with imposing restrictions on sentence recommendations “would disappear if the sentence could be manipulated through charge bargaining”); Bibas, Outside the Shadow, supra note 17, at 2536 (“As long as prosecutors can manipulate baseline charges, trying to cap discounts is hopeless.”). Covey contends that his proposed plea ceilings (which cap increases in post-trial sentences as opposed to decreases in post-plea sentences) would be “harder for prosecutors to evade through charge . . . bargaining,” Covey, supra note 179, at 1273, but his argument seems to require the absence or blanket invalidation of mandatory minimums. See id. at 1276 n.176.

Charge sliding, however, can instead be regulated directly. Indeed, much as the ability to combine charges together is a direct function of the law of joinder, the prosecutor’s ability to trade charges for guilty pleas is a direct function of the law of amendment and dismissal, which together determine when and whether one set of charges can be replaced with another—a necessary component of any bargain. 186 “Amendment” refers to replacing one charge with another, whereas “dismissal” (sometimes called “nolle prosequi”) refers to eliminating a charge without offering a replacement. Specifically, procedural law can regulate charge sliding in four basic ways.

Figure 17: Conceptual Approaches to the Law of Charge Amendment and Dismissal


In the least restrictive of these models (the leftmost above), the prosecutor’s ability to swap or reduce her charges is simply unlimited, affording her maximal charge-sliding flexibility. At the other end of the spectrum, by contrast, amendments and dismissals of charges are simply prohibited once the charges are filed—an approach that would essentially prohibit charge bargaining altogether, much as Wright and Miller propose. 187 See Wright & Miller, supra note 10, at 49. But cf. infra note 211 (discussing “precharge bargaining”). Between these poles, however, two intermediate models cabin charge sliding without eliminating it. The first gives judges discretion to reject proposed changes to the charging instrument, either by requiring judicial approval for amendment and dismissal in general or by expressly authorizing judges to reject plea agreements that entail charge reductions the judge deems inappropriate. 188 The latter option can serve as a nudge to judges, insofar as it attaches the power to oversee charge dismissals and amendments to the practice of charge bargaining itself. See, e.g., Fed. R. Crim. P. 11(c) (“[A] plea agreement may specify that an attorney for the government will . . . not bring, or will move to dismiss, other charges . . . . [T]he court may accept the agreement, reject it, or defer a decision until the court has reviewed the presentence report.”); see also United States v. Walker, No. 2:17-CR-00010, 2017 WL 2766452, at *1 (S.D.W. Va. June 26, 2017) (holding that in applying Rule 11, “[i]t is the court’s function to prevent the transfer of criminal adjudications from the public arena to the prosecutor’s office”); id. (rejecting a plea agreement that would have slid down from an indictment charging five counts of drug distribution and one count of being a felon in possession of a firearm to “a single count of possession with intent to distribute”). Alternatively, rather than relying solely on individual judges’ discretion, the procedural regime could instead place firm caps on charge sliding by permitting amendment only if the initially charged offense is replaced by one only modestly less severe. 189 The approach here is similar to Gazal-Ayal’s proposed cap on sentence discounts, see supra note 184, with the key difference that it regulates charge reductions directly.

And once again, states resolve this procedural issue in different ways.

Figure 18: Variability Across States in the Law of Amendment and Dismissal 190 For sources, see infra Appendix at Table E.


Thus, while thirteen states afford prosecutors the full range of flexibility in charge sliding, the majority (thirty-three) require judicial approval before charges can be amended or dismissed—with fifteen of those expressly granting judges the authority to reject charge bargains that they deem inappropriate. Four other states, however, impose additional re­straints: New York has essentially enacted a charge-bargaining variant of Gazal-Ayal’s proposal, barring prosecutors from dismissing charges unless they replace them with charges from a proximate tier of offense, 191 See, e.g., N.Y. Crim. Proc. Law § 220.10(5)(a)(iii) (McKinney 2007) (“Where the indictment charges one of the class B felonies . . . any plea of guilty . . . must be or must include at least a plea of guilty of a class D felony.”); id. § 220.10(5)(a)–(d) (prescribing additional tier requirements). Note that the New York model applies only to indictments, creating an opportunity for prosecutors to evade the restriction by engaging in charge bargaining before formal charges are filed. For further discussion of “precharge bargaining,” see infra note 211. while Michigan has adopted a similar regime for certain drug offenses. 192 See infra note 338 (describing the Michigan regime). Meanwhile, California and Nevada impose bans on charge sliding for certain classes of offenses. 193 Notably, because these bans apply only in certain cases, the regimes are more restrictive than New York’s capped regime in those instances when the ban applies, but less restrictive in those when it does not. See Cal. Penal Code § 1192.7 (2018) (banning plea bargaining for serious offenses and certain defendants); Nev. Rev. Stat. §§ 484C.430, 484C.440, 484C.470 (2015) (banning bargaining for DUI-related offenses). In each of these states, procedural law enacts regulatory constraints that scholars hoping to restrict charge sliding have proposed doing less directly.

B. Verdict Sliding: The Law of Lesser Offenses

Finally, while the law of amendment and dismissal can regulate charge sliding directly, the law of lesser offenses can do so indirectly, by removing some of prosecutors’ incentives to inflate charges at the outset of the case. To appreciate how this body of law might influence plea bargaining dynamics, imagine (for the sake of simplicity) that our hypothetical armed robbery defendant is charged with only a single, inflated count of kidnapping. Assume, moreover, that if the case were to go to trial, the jury would recognize the kidnapping charge for the overreach that it is—but would be comfortable convicting the defendant on some less serious offense (for example, robbery). If kidnapping is the only charge on the table, the jury will be in a bind: It can choose to punish the prosecutor’s overreach with an acquittal, or, reluctant to let someone whom it believes to be an armed robber go free, it can hold its nose and convict the defendant of kidnapping. By contrast, if at the end of the trial the jury is given the option to convict the defendant of robbery as a lesser offense (even though the only charge initially filed was kidnapping), then the pressure will be off: It can and presumably will slide down to the robbery verdict.

Here, however, is the wrinkle: When the prosecutor and the defendant face each other across the bargaining table—well in advance of any trial—neither will be in a good position to predict how a future jury might resolve the dilemma just described. For not only will they have no idea who the jurors are, they will also have only a dim sense of how the evidence will unfold in the heat of trial and of how the jury will react to it. Given this ex ante ambiguity, fixed rules either requiring the jury to consider lesser offenses or prohibiting them from doing so will be unlikely to have much effect on plea bargaining dynamics—because neither party will know whose ox will be gored in the end. Similarly, a rule that affords each party the right to put a lesser offense before the jury will not have much impact either, as both parties’ willingness to risk going to trial will be increased by the knowledge that if he or she wants the jury to have access to the “soft-landing” of a lesser-offense verdict, that option will be available. 194 Cf. Catherine L. Carpenter, The All-or-Nothing Doctrine in Criminal Cases: Independent Trial Strategy or Gamesmanship Gone Awry?, 26 Am. J. Crim. L. 257, 277 (1999) (“Originally designed as a tool to aid the prosecution[,] . . . [lesser-offense instructions were] quickly recognized as a valuable aid for the defendant [insofar as they] afforded the jury a less drastic alternative than the choice between conviction of the offense charged or acquittal.”). In other words, so long as juries are not systematically likely to break one way or the other when faced with the dilemma of inflated charges, symmetrical rules governing the availability of lesser-offense instructions will in most cases be a wash. 195 The subconstitutional law of lesser offenses is thick with different fixed ex ante rules that resolve issues ranging from what counts as a “lesser offense” in the first place, to the amount of evidence that must support such a charge before it can be put to the jury, to what order the jurors must follow when considering the various charges at hand, to whether or not the jury must be unanimous in its verdict. See id. at 265–72 (discussing different approaches to defining “lesser offense” and different evidentiary standards for issuing lesser-offense instructions); see also 6 LaFave et al., supra note 73, § 24.8(e), at 637–48 (same); cf. Blueford v. Arkansas, 566 U.S. 599, 619–20 (2012) (Sotomayor, J., dissenting) (discussing the indeterminate beneficiaries of acquittal-first rules). Taken together, these features of the subconstitutional law can substantially impact how easy or hard it will be for a jury to slide down from an overplayed hand in the context of any given case. But because the desirability of such sliding down will be hard for each party to predict (and thus to optimize) ex ante, it is difficult to assess the regulatory shadow that these rules might cast on the bargaining table in the typical case.
For a crisp illustration of this point, consider a recent procedural debate in Oregon, where, according to the Wall Street Journal, the Oregon District Attorney’s Association promoted “a state constitutional amendment that could make it harder for [prosecutors] to win convictions.” Jacob Gershman, Oregon Prosecutors Push to Make Some Convictions Harder to Win, Wall St. J. (Jan. 12, 2018),
articles/oregon-prosecutors-push-to-make-some-convictions-harder-to-win-1515797099 (on file with the Columbia Law Review). In truth, the proposed amendment was more complicated than that description suggests, as the amendment would have repealed an existing “constitutional provision that makes it possible for criminal defendants to be found guilty or acquitted” by a nonunanimous jury. Id. In other words, the proposed amendment would have repealed a procedural rule that “helps to avoid a hung jury” and that thus “contributes to more cases being resolved in both directions,” including—in some hard-to-predict number of cases—by conviction. Id. (emphasis added) (internal quotation marks omitted) (quoting Josh Marquis, Clatsop County District Attorney). Given the ambiguous ex ante effects of the current rule (and of the proposed change), it is not surprising that “support for a repeal [wasn’t] unanimous,” even “[a]mong Oregon’s local prosecutors.” Id. Indeed, a spokesperson for the Oregon Criminal Defense Lawyers Association initially supported the proposed change, predicting that a move to do “away with non-unanimous juries would have gained widespread support from an array of stakeholders.” Katie Shepherd, Oregon District Attorneys Drop Plan to Scrap Non-Unanimous Jury Verdicts, Willamette Week (Jan. 30, 2018), [http://]. In the end, however, the prosecutors tied their unanimous-jury proposal to a separate proposed reform that would have clearly benefitted prosecutors over defendants: They pushed to repeal another state constitutional provision that currently affords “defendants the right to choose a bench trial in front of a judge over a jury trial without the prosecution’s consent.” Id. (emphasis added). Once the symmetrical reform regarding jury unanimity was yoked to an asymmetrical reform uniquely benefitting prosecutors (insofar as it sought to repeal an asymmetrical existing rule uniquely benefitting defendants), “law professors, criminal justice reform advocates, and the Oregon Criminal Defense Lawyers Association” pulled their support, which in turn prompted the prosecutors to abandon both efforts. Id.

Asymmetrical rules, by contrast, can have a very different effect. For if the prosecutor knows at the outset that come the end of trial she alone will get to decide whether the jury is instructed on lesser offenses, then she will have every incentive to overcharge at the beginning of the case: She can choose her charges confident in the knowledge that the jury will have access to a “soft-landing” verdict only if she wants it to—and that it will face an all-or-nothing choice if she deems that to be most advantageous. Conversely, if the law lets the defendant unilaterally decide whether to give the jury a lesser-offense instruction, then the prosecutor will know up front that she will always confront the worst-case scenario at the end of the trial: She will be denied a soft landing whenever her adversary thinks she needs it, but he will be able to benefit from his own soft landing if he thinks that is in his best interest.

Given these dynamics, the law of lesser offenses can impact plea bargaining dynamics in three basic ways.

Figure 19: Variability Across States in the Law of Party Control over Lesser-Offense Option 196 For sources, see infra note 197.


Specifically, as the preceding discussion sets out, asymmetrical rules—occupying here the poles of the spectrum above—can afford the prosecutor (on the left) or the defendant (on the right) the ability to enter plea negotiations confident that, come the end of trial, she or he will have an advantage when it comes to determining the choice architecture for the jury’s deliberations. And in fact, a handful of states embrace asym­metrical rules in this domain: Texas makes it easier for prosecutors to obtain lesser-offense instructions while Colorado, New Jersey, and Utah give defendants greater control over how juries will evaluate the charges against them. 197 For Texas’s prosecution-favoring rule, see Grey v. State, 298 S.W.3d 644, 645 (Tex. Crim. App. 2009) (holding that when the defendant requests a lesser-offense instruction, “some evidence must exist in the record that would permit a jury rationally to find that . . . he is guilty only of the lesser offense” but that the prosecutor is not similarly restricted (emphasis added) (internal quotation marks omitted) (quoting Rousseau v. State, 855 S.W.2d 666, 673 (Tex. Crim. App. 1993))). For defendant-favoring rules, see People v. Skinner, 825 P.2d 1045, 1047 (Colo. App. 1991) (“[A] lesser non-included offense instruction may be given only if the defendant requests it or consents to it.”); State v. Thomas, 900 A.2d 797, 804 (N.J. 2006) (restricting “prosecution requests for included offense jury charges solely to those offenses that are lesser-included offenses within the meaning of [a specified statute],” while “[i]n contrast” permitting the defendant to obtain an instruction on any lesser charge for which “the evidence provides a rational basis”); State v. Baker, 671 P.2d 152, 156–58 (Utah 1983) (instructing that defense-requested lesser-offense instructions should be given whenever the evidence provides a rational basis for a verdict on the offense, while limiting prosecution-requested instructions to those with elements subsumed in the original charge). In addition to the asymmetrical rule favoring defendants in Colorado set out in Skinner, Colorado also has a statute that allows the defendant to ask the court to make the prosecutor choose which charge to send to the jury whenever multiple charges “are supported by identical evidence.” Colo. Rev. Stat. § 18-1-408(3) (2017). Beyond these three states, Arizona gives defendants an asymmetrical right to force an all-or-none strategy in capital cases. See State v. Rodriguez, 921 P.2d 643, 652 (Ariz. 1996) (“If [the defendant] objects, the instruction should not be given.”); State v. Krone, 897 P.2d 621, 625 (Ariz. 1995) (“A defendant should not have a lesser included instruction forced upon him.”). But cf. State v. Gipson, 277 P.3d 189, 190–91 (Ariz. 2012) (rejecting such a right for noncapital defendants). In these four states, this procedural asymmetry casts its shadow back onto the plea bargaining table and thereby indirectly restricts or enhances—that is to say, regulates—prosecutorial power.

V. The Hydraulics and the Politics of Reform

To this point, this Article has mapped a series of regulatory levers that reside within a long-overlooked body of subconstitutional law, demon­strating for each not only its potential to facilitate or restrict prosecutorial power but also its heterogeneous deployment across jurisdictions. 198 As explained supra note 22, this Article uses charge bargaining as its primary lens because charge bargaining is widely seen as the central instrument of prosecutorial power. Other such mechanisms exist, however, and the ways in which they are regulated by subconstitutional procedural law is ripe for future study. Two examples of such mechanisms include:
(1) Sentence bargaining, whereby the prosecutor negotiates with the defendant over the sentence that she will recommend to the judge if the defendant pleads guilty. While the judge’s involvement generally renders this practice less prone to abuse than charge bargaining, prosecutors can still obtain undue leverage when engaging in sentence bargaining by exploiting defendants’ uncertainty about judges’ future sentencing decisions. See, e.g., Bibas, Outside the Shadow, supra note 17, at 2533 (discussing challenges facing defendants when trying to anticipate how judges will sentence). Prosecutors may also exploit judges’ comparative lack of access to facts about the underlying criminal conduct, which can cause judges to defer too much to prosecutors’ sentencing recommendations. See, e.g., Berthoff v. United States, 140 F. Supp. 2d 50, 61–67 (D. Mass. 2001) (discussing challenges facing judges in imposing a sentence following a guilty plea). Subconstitutional procedural law, however, can help on both fronts. With respect to defendants’ uncertainty, procedural law might authorize judges to give defendants clarity by allowing judges to declare up front (that is, before plea negotiations conclude) the sentence they are likely to impose; alternatively, procedural law could permit the parties to submit an agreed-upon sentence to the judge for approval, an approach that would offer both clarity and judicial oversight at the same time. And in fact, states already experiment with both approaches. See Wright & Miller, supra note 10, at 89 (“To give defendants more complete and reliable information, a growing number of states encourage rather than forbid judicial involvement in plea discussions.”); see also Idaho R. Crim. P. 11(f)(3) (“If the court accepts the plea agreement, . . . it will be bound by the terms of the plea agreement in the final disposition of the case.”). As for undue judicial deference to prosecutorial sentencing recommendations, procedural law might require that detailed presentence reports be prepared by judicial personnel, see, e.g., Mich. Ct. R. 6.425, while perhaps also limiting—or even eliminating—the prosecutor’s ability to recommend a specific sentence, cf. Ariz. R. Crim. P. 26.10; Ind. Code §§ 35-38-1-3, -11 (2017); Me. R. Crim. P. 32(a)(2); Nev. Rev. Stat. § 176.015 (2015). For examples of how asymmetrical sentence-recommendation rules might operate in practice, see Bibas, Outside the Shadow, supra note 17, at 2518 (noting that “[i]n about one-third of all federal districts, federal prosecutors follow a policy of not recommending specific sentences for cooperating defendants”); Wright & Miller, supra note 10, at 80 (describing a practice in New Orleans whereby “most of the judges declared that they discourage prosecutors from expressing any opinion at all about the proper sentence”).
(2) As I have noted in prior work, another mechanism of prosecutorial power arises from the fact that prosecutors, as repeat players, often care about obtaining favorable legal precedents more than defendants do, and are thus in a position to engage in claim bargaining. See Crespo, Regaining Perspective, supra note 13, at 2030–38. A claim bargain occurs when prosecutors deploy their charge- or sentence-bargaining leverage not to secure a guilty plea but rather to get the defendant to forfeit a potentially viable legal claim—a power prosecutors can use to shape the underlying substantive and constitutional landscapes themselves. See id. at 2030–35. The prosecutor’s ability to deploy this leverage, however, depends in part on whether the defendant has any opportunity under the governing procedural law to separate the litigation of any such legal issues from negotiations over his factual guilt or innocence. See id. at 2033–34. And on that score, procedural law varies. For example, in some states suppression motions must be entertained in the course of the defendant’s preliminary hearing, while in others they are deferred until trial—by which point a plea agreement may require the defendant to waive the issue. Compare, e.g., 725 Ill. Comp. Stat. Ann. 5/109-3(e) (West 2006) (allowing for resolution of suppression issues at any preliminary hearing), Mich. Ct. R. 6.110(D)(2) (same), and Tenn. R. Crim. P. 5.1(a)(1) (same), with, e.g., Md. Rule 4-221 (“[E]vidence may not be excluded [at a preliminary hearing] on the ground that it was acquired by unlawful means.”). Similarly, the law governing conditional plea agreements can determine whether the defendant will have a chance to acknowledge his factual culpability (and thus plead guilty) while simultaneously preserving his opportunity to challenge and appeal potentially dispositive legal issues. See Crespo, Regaining Perspective, supra note 13, at 2037–42 (discussing ways to regulate “claim bargaining” by amending conditional-plea rules and appeal-waiver rules); cf. Class v. United States, 138 S. Ct. 798, 803–805 (2018) (establishing a default rule against reading a guilty plea as an appellate waiver of certain legal claims).
That heterogeneity, in turn, suggests a genuine opportunity for regulatory experimentation, and with it the potential for reform. And yet, as the discussion also shows, assessing the practical impact these levers might hold is complicated, particularly if they are thrown haphazardly or in isolation—for as Professor Bowers observes, “[p]lea bargaining has something of a hydraulic quality” to it that affords it “a remarkable resistance to change.” 199 Josh Bowers, Response, The Unusual Man in the Usual Place, 157 U. Pa. L. Rev. PENNumbra 260, 269 n.39 (2009), (on file with the Columbia Law Review). Predicting the ultimate impact of law reform in this arena thus requires not only excavating and mapping plea bargaining’s hidden law—the project undertaken thus far—but also situating that newly excavated regulatory framework within its surrounding sociolegal dynamics, and analyzing their interrelation.

That is a research agenda capacious beyond one article’s ability to satisfy. It is an agenda, however, that this project both invites and begins to equip scholars to undertake. The goal in this Part is thus to identify a core set of questions and hypotheses for future investigation. The first of these questions concerns the familiar relationship between law’s form and its function, asking how scholars might best draw on the insights offered here to better assess the impact that procedural reforms might have on plea bargaining practices on the ground. The second question then turns to plea bargaining’s political economy, asking which institutional actors are most empowered to promote or prevent these reforms, and how their motivations might factor into the analysis.

A. The Hydraulics of Reform: Assessing Hidden Law’s Potential Impact

When it comes to assessing the potential impact of procedurally grounded reforms, it is useful to consider potential hydraulic counter­pressures that are both internal and external to the legal apparatus itself.

1. Internal Hydraulics: The Importance of Coordinated Regulatory Frameworks. — With respect to counterpressures internal to law, the discussion above has identified multiple instances in which efforts to clamp down on one prosecutorial tactic (by reconfiguring one particular legal lever) could simply divert prosecutors to some other tactic that the law regulates less stringently. 200 See, e.g., supra notes 60–61 and accompanying text (discussing the hydraulic interrelationship of joinder and preclusion); supra note 84 and accompanying text (discussing the interrelationship of joinder and cumulative sentencing); supra notes 156–159 and accompanying text (discussing the interrelationship of factual and legal overreaching). Of course, plea bargaining is not unique in this respect: Regulated actors routinely seek to evade their constraints by exploiting potential cracks in the regulatory regime, an endemic feature of regulation that simply underscores the need to avoid such cracks in the first place—by constructing a coordinated and up-to-date regulatory apparatus.

Fortunately for plea bargaining reformists, the analysis in the preceding Parts helps identify precisely such a coordinated set of reforms, insofar as it highlights how different procedural levers interact with one another and offers real-world examples of cutting-edge procedures already in place across the states. Drawing on that analysis, the following three hypothetical rules offer an example of a holistic regulatory approach that would constrain prosecutorial power by pulling on each of the various procedural levers discussed thus far:

Rule 1Joinder, Preclusion, and Sentencing: For any given criminal event, episode, or scheme, the prosecution may file only one case against each defendant and, in that case, may charge only one offense, unless the prosecution alleges that the defendant caused serious bodily injury to more than one person, in which case the prosecution may join together one charge per each alleged victim. 201 See supra sections II.A– .B; cf. Ark. Code Ann. § 16-85-404 (1987) (establishing a rule of nonjoinder, with an exception when “[t]he homicide of several persons [is allegedly] committed by the same person”), repealed by Arkansas Criminal Code Revision Commission’s Bill, Act 1994, § 503, 2005 Ark. Acts 6932, 7459–61; N.C. Gen. Stat. § 15A-926 (2017) (common-scheme preclusion); Tex. Penal Code § 3.04 (2017) (universal severance). In the event multiple charges yield multiple convictions in a single case, the associated penalties shall be served concurrently, unless the court orders otherwise. 202 See supra section II.C; cf. Mo. Sup. Ct. R. 29.09 (default concurrent sentences).

Rule 2Screening and Dismissal: The defendant shall have a right to a preliminary hearing at which the prosecutor must demonstrate via evidence admissible at trial that the charges filed are supported by probable cause. 203 See supra section III.A; cf. Tenn. R. Crim. P. 5(e)(4) (hearing as of right). The court shall dismiss any charges that are unsupported by the evidence adduced at such hearing or that fail to state an offense as a matter of law. 204 See supra sections III.A–.B. The court shall also have the authority to dismiss charges when the interests of justice otherwise so require. 205 See supra section III.C; cf. Vt. R. Crim. P. 48(b)(2) (equitable dismissal). In the event of dismissal, no charges arising from the same alleged criminal event, episode, or scheme may subsequently be filed against the defendant, unless supported by newly discovered evidence that was not previously available to the prosecution. 206 See supra section III.A.3; cf. R.I. R. Crim. P. 5(c) (no refiling absent new evidence).

Rule 3Amendment and Lesser Offenses: Prior to trial, charges may be amended or dismissed only by leave of court, and only when the proposed new charges carry sentencing consequences propor­tionate to the original charges. 207 See supra section IV.A; cf. Ark. R. Crim. P. 25.3 (express charge-bargaining review); N.Y. Crim. Proc. Law § 220.10(5)(a)(iii) (McKinney 2014) (capped charge discounts). Once trial has commenced, the jury will not be instructed with respect to any charges other than those in the charging instrument, unless the defendant so requests. 208 See supra section IV.B; cf. State v. Krone, 897 P.2d 621, 625 (Ariz. 1995) (affording defendant the asymmetric option to obtain a lesser-offense instruction in capital cases); McConnell v. State, 110 N.W. 666, 667 (Neb. 1906) (affording defendant a similar option in all cases), overruled by State v. Pribil, 395 N.W.2d 543, 549 (Neb. 1986).

By yoking together a one-charge-per-crime rule, a robust set of screening mechanisms, and judicially managed restrictions on downward bargaining, this hypothetical suite of reforms would dampen prosecutors’ plea bargaining leverage across multiple fronts, ultimately making prosecution (and with it imprisonment) a more costly affair. That is an unsurprising feature of a reform agenda that hopes to counter mass incarceration. 209 See supra notes 17–18 and accompanying text (discussing critiques of American incarceration rates). And yet, the potentially deterrent or incapacitative upsides of incarceration will be diminished by these reforms as well. Reasonable minds thus can and should disagree about whether the particular reforms suggested above or others like them go too far—or not far enough. 210 Compare Randall Kennedy, Race, Crime, and the Law 305–06 (1st ed. 1997) (cautioning against criminal justice policies that are “inattentive to the aspirations, frustrations, and fears of law-abiding people compelled by circumstances to live in close proximity to [criminality, including] . . . [that] sector of the black law-abiding population that desires more rather than less prosecution and punishment for all types of crime”), with Allegra M. McLeod, Prison Abolition and Grounded Justice, 62 UCLA L. Rev. 1156, 1161 (2015) (offering a “sustained discussion of . . . a ‘prison abolitionist framework’ [that is] oriented toward substituting a constellation of other regulatory and social projects for criminal law enforcement”). Cf. Herbert L. Packer, The Limits of the Criminal Sanction 153 (1968) (discussing inherent tension between liberty and security in the administration of criminal justice). One might expect the ideal answer to this policy question (or the range of acceptable answers) to vary from one locality to the next, given the often-sharp variation in empirical realities, political challenges, and normative approaches at play in criminal justice systems across the country. On the importance of both disaggregation and localism in criminal justice administration, see, for example, Lanni, supra note 28, at 387–90; Sara Mayeux, The Idea of “the Criminal Justice System,” Am. J. Crim. L. (forthcoming 2018) (manuscript at 6–9), (on file with the Columbia Law Review); William J. Stuntz, Unequal Justice, 121 Harv. L. Rev. 1969, 1974 (2008). Indeed, that is the core normative question latent within plea bargaining’s hidden regulatory framework that this project aims to surface, albeit not to resolve. But regardless of how policymakers ultimately strike that balance, the hypothetical rules set out above demonstrate that, simply as a matter of regulatory design, a holistic approach to plea bargaining reform—grounded in subconstitutional procedural law—is conceptually within reach. 211 For a discussion of the identity and the potential motivations of the policymakers responsible for this body of law, see infra section V.B. As for prosecutorial evasion of procedural restraints, two additional points merit further discussion:
First, there is the concern that prosecutors will bargain around the enhanced procedures themselves. Reforms creating a robust preliminary hearing, for example, might arguably be undercut if intrepid prosecutors use their charge-bargaining leverage to force defendants to waive the hearing itself. Cf. 4 LaFave et al., supra note 73, § 14.2(e), at 349 (noting “a prosecution practice of offering significant concessions to defendants who waive their preliminary hearings”); Gazal-Ayal, supra note 34, at 2331 (“Prosecutors can, and often do, condition exceedingly lenient bargains on a waiver of preliminary hearing or grand jury indictment.”). On closer inspection, however, such “procedure bargaining” emerges not so much as regulatory evasion but rather as regulation in action, because enhanced procedural protections ought to raise the “price” that prosecutors must pay for a guilty plea in the first place. Thus, a prosecutor who wants to shield her complaining witness from adversarial pretrial cross-examination will have to offer a defendant a markedly more favorable plea deal (with less prison time) in a jurisdiction where the defendant has a procedural right to force the complainant to testify at a preliminary hearing than in a jurisdiction where a police officer can supply all of the relevant evidence via hearsay—or in which preliminary hearings simply don’t exist at all. Cf. supra section III.A.2 (discussing application of hearsay rules to preliminary hearings). Plea bargains, in other words, take place in the shadow of the procedural law, with each procedural lever constituting a stick in the defendant’s bundle of rights that he can use as his own form of leverage to counterbalance the prosecutor’s leverage at the bargaining table. Cf. Robert H. Mnookin & Lewis Kornhauser, Bargaining in the Shadow of the Law: The Case of Divorce, 88 Yale L.J. 950, 966–73 (1979) (articulating a model of bargaining in which “legal rules . . . give[] each [party] certain bargaining chips”); John Rappaport, Unbundling Criminal Trial Rights, 82 U. Chi. L. Rev. 181, 181–82 (2015) (suggesting defendants might benefit from being able to “‘unbundle’ their jury trial rights and trade them piecemeal . . . to reduce their sentencing exposure”).
More fundamentally, the prosecutor’s ability to engage in procedure bargaining is itself determined by subconstitutional procedural law, which determines in the first instance which procedural rights are waivable, and thus negotiable. Cf. State v. Ethington, 592 P.2d 768, 769–70 (Ariz. 1979) (deeming the right to appeal nonnegotiable). That question, moreover, must be resolved one way or the other for virtually every regulatory lever discussed in this Article, which raises the possibility that certain forms of procedure bargaining might simply be banned outright. How such a ban might ultimately impact the underlying market for pleas will be context dependent, driven in part by prosecutorial resources. The implications—and the wisdom—of such a ban is thus open to debate. Cf. Crespo, Regaining Perspective, supra note 13, at 2039 (proposing a procedural rule that would render certain constitutional claims nonwaivable); Rappaport, supra, at 196 (arguing against such inalienability, while acknowledging that “[s]ocial harms may . . . justify restraints on the alienation of [some procedural] rights”). The key point here, however, is that this policy debate is one for subconstitutional procedural law to resolve.
Second, apart from procedure bargaining, there is a related question about whether prosecutors might seek to evade procedural constraints by moving plea bargaining earlier in time, before the judicial process and its accompanying regulatory framework gets into gear. Cf. Gazal-Ayal, supra note 34, at 2342 (“The concern that prosecutors would find ways to continue to charge bargain on the sly always exists whether the practice is totally prohibited or only restricted and subjected to courts’ review.”); Wright & Miller, supra note 10, at 47 n.60 (“Any effort to ban or limit plea bargains must account for the possibility of precharge bargaining.”). Notably, however, this concern is considerably more pronounced in the federal system, in which plea bargaining can occur during the lengthy investigation periods that precede the commencement of formal proceedings. In state courts, by contrast, the vast majority of criminal cases are initiated by an arrest and thus come into contact with the formal judicial system either in conjunction with or soon after the prosecutor’s first engagement with the case. See supra note 107 (describing Gerstein hearings). Moreover, plea bargaining cannot occur until the defendant has an attorney. See Padilla v. Kentucky, 559 U.S. 356, 373–74 (2010) (holding that the “negotiation of a plea bargain is a critical phase of litigation for purposes of the Sixth Amendment right to effective assistance of counsel”). And because the vast majority of state court defendants are indigent, their attorneys are often appointed through the court system itself, which further confines the plea bargaining process to a point in time after the judicial process—and its attendant regulatory procedures—kicks into gear. Thus, the practical window within which precharge bargaining and any associated regulatory evasion might occur is substantially shorter in state court than in the federal system.

2. External Hydraulics: The Importance of Studying Procedural Law’s Potential Regulatory Impact. — There remains, however, a separate question: To what extent might hydraulic forces external to law stymie even a carefully calibrated set of regulatory interventions? Few scholars of American criminal justice doubt that such extralegal forces—ranging from resource imbalances between prosecutors and defendants, to informal institutional norms and practices, to the complex power dynamics associ­ated with race, gender, and class—produce sometimes-sizable gaps between the criminal law codified on the books and the criminal law implemented on the ground. 212 See, e.g., Issa Kohler-Hausmann, Jumping Bunnies and Legal Rules: The Organizational Sociologist and the Legal Scholar Should Be Friends, in The New Criminal Justice Thinking 246, 246 (Sharon Dolovich & Alexandra Natapoff eds., 2017) (“[T]he fact that law in action does not match law on the books is as predictable as the fact that a windup jumping bunny toy will jump after being wound up.”); Alexandra Natapoff, The Penal Pyramid, in The New Criminal Justice Thinking, supra, at 71, 71–73, 76, 88 (theorizing a “penal pyramid” in which the “explanatory power” of legal rules “ebbs and flows” vis-à-vis extralegal forces); id. at 72–73 (arguing that formal law “govern[s] practices and outcomes” in “the world of federal offenses, serious cases, and well-resourced defendants,” but may “hardly matter at all” in the resolution of “petty” misdemeanors, which “number in the thousands”). On the significance of resource imbalances between prosecutors and defense attorneys, see, for example, Crespo, Regaining Perspective, supra note 13, at 2019–20 (discussing the indigent-defense crisis); see also Bibas, Outside the Shadow, supra note 17, at 2539–40 (discussing consequences for plea bargaining). But cf. Natapoff, supra, at 75 (observing that “in some jurisdictions an elite public defense bar provides stellar representation to the very poorest black defendants”). On the significance of informal norms and practices, see, for example, Mona Lynch, Hard Bargains: The Coercive Power of Drug Laws in Federal Court 2–8 (2016) [hereinafter Lynch, Hard Bargains]. As for race, class, and gender dynamics, the literature is simply too massive to cite. The persistent challenge is sorting out how the formal law that creates and potentially regulates prosecutors’ charge-bargaining leverage interacts with these broader forces to impact plea bargaining practices on the ground. 213 As Professor Alexandra Natapoff observes, criminal procedure scholars may sometimes appear “to talk past each other,” insofar as “[s]ome work focuses on the rules” codified in formal legal texts, while other work focuses on “the power dynamics” driving criminal law’s administration. Natapoff, supra note 212, at 74. This project aims to straddle those two conversations: It joins those scholars who see plea bargaining’s pathology as arising from an overly broad substantive criminal law and an overly permissive constitutional law of criminal procedure, see supra notes 5–6, 27–30; it urges a broader conception of the legal frameworks that influence plea bargaining; and it ultimately aims to expand the field of potential law reform. In other words, while this Article raises the question of which law matters, it proceeds from a shared scholarly premise that law does indeed matter in this domain. But it also acknowledges the limitations of a law-centric account, particularly when it comes to state court adjudication of felony cases—the central focus of this Article—which Natapoff aptly describes as the criminal justice system’s “murky middle,” where a “combination of legal rules and social facts [together] generate outcomes that cannot be fully explained by either alone.” Natapoff, supra note 212, at 73. In short, the analysis here proceeds from the premise, articulated by Professor Issa Kohler-Hausmann, that “legal rules make out a set of formal directives outlining how things ought to operate, who has which powers, and what activities are authorized under what conditions,” which legal actors “must then operationalize . . . in order to do their work.” Kohler-Hausmann, supra note 212, at 252. In so doing, the project sets the stage for further study of how these newly uncovered legal frameworks “influence organizational and institutional dynamics within their respective fields of operation.” Id. at 260.

In a criminal justice system as atomized as ours, there can be no single answer to that question—only a plethora of very specific ones, each tied to the particular sociolegal facts of the jurisdiction at hand. 214 Cf. Mayeux, supra note 210 (manuscript at 3) (writing against “the premise that there is, in fact, some ‘system’” to be found amidst American criminal law’s “tens of thousands of functionally related though formally distinct entities of an almost impossibly wide-ranging set of sizes, scales, aims, and types”). Surfacing the formal legal variation among those jurisdictions, however, is an important first step in examining how institutional actors within each one “make sense” of the governing legal rules and “decide what they mean in particular action settings.” 215 Kohler-Hausmann, supra note 212, at 266–67 (observing that an integrated view of the explanatory power held by both formal law and extralegal forces “generates a set of research questions about how legal rules are fundamentally always interpolated into the course of ongoing activity”). Indeed, to the quantitative empiricist, the discussion in the preceding Parts may offer some early grist for a study of the consequences that one should expect to follow from pulling one or another of the regulatory levers uncovered here, as each lever is essentially an independent variable of interest, paired with a hypothesis about how it ought to impact plea bargaining power and criminal justice outcomes. 216 To be sure, myriad confounding variables are also at play. See infra note 222 and accompanying text. And accessing criminal justice data presents its own challenges. See Crespo, Systemic Facts, supra note 6, at 2108–10 (discussing the lack of data transparency in many criminal justice systems). Consider, in this vein, the following two matrices, which collect some of the key variables that one might hope to explore:

Table 2: Composite Procedural Levers of Individual States, Ranked by Plea Rates 217 For sources, see infra Appendix.

State Piling On Overreaching Sliding Down Plea Rate Dismissal Rate Trial Rate
Joinder–Severance Preclusion Cumulative Sentencing Pretrial Evidentiary Review Amendment
NY 86.9% 6.7% 4.1%
MO 83.9% 10.8% 1.5%
AZ 81.8% 15.8% 2.4%
CA   80.9% 16.7% 2.3%
KS 80.1% 14.4% 2.8%
TX 77.2% 20.4% 2.4%
IN 77.0% 17.9% 2.1%
VT 76.6% 20.4% 1.6%
FL 76.3% 8.3% 1.9%
AK 75.4% 20.5% 3.6%
WI 74.7% 19.7% 2.6%
NM 73.8% 23.9% 1.8%
68.3% 14.2% 14.3%
NC 66.9% 31.5% 1.7%
HI 66.3% 15.0% 5.5%
KY 62.3% 33.7% 4.0%
IL 61.7% 33.8% 4.6%
TN 55.7% 39.3% 1.8%

Broadly Restrictive Regimes (Black); Nonrestrictive Regimes (White); Intermediate Regimes (Gray).

Table 3: Composite Procedural Levers, Ranked by Felony Pleas per 100 Crimes 218 For sources, see infra Appendix.

State Piling On Overreaching Sliding Down Felony Pleas per 100 Crimes Plea Rate Dismissal Rate
Joinder–Severance Preclusion Cumulative Sentencing Pretrial Evidentiary Review Amendment
NC* 31 66.9% 31.5%
IN 27 77.0% 17.9%
NM 21 73.8% 23.9%
FL 20 76.3% 8.3%
MO 19 83.9% 10.8%
VT 18 76.6% 20.4%
AK 18 75.4% 20.5%
WI 18 74.7% 19.7%
CA 17 80.9% 16.7%
KY* 17 62.3% 33.7%
TN* 17 55.7% 39.3%
TX 16 77.2% 20.4%
AZ 15 81.8% 15.8%
KS 15 80.1% 14.4%
IL* 15 61.7% 33.8%
14 68.3% 14.2%
NY 10 86.9% 6.7%
HI* 4 66.3% 15.0%

Broadly Restrictive Regimes (Black); Nonrestrictive Regimes (White); Intermediate Regimes (Gray). Asterisks in Table 3 mark the states with the five lowest plea rates in Table 2.

Taken together, these tables capture a promising degree of variability with respect to the essential data points of interest. Specifically, the left-hand portion of each table captures the key independent variables, show­casing genuine variability with respect to each state’s overarching regulatory regime. It does so by presenting a composite sketch of each state’s procedural law, with the various columns corresponding to the main procedural levers discussed throughout this Article. 219 For ease of reference, the tables focus on the main levers discussed in Parts II–IV, though they could easily be expanded to include the full array. Note also that both tables are limited to states in which reliable plea bargaining data are available. For sources, see infra Appendix at Table F. If a state employs a nonrestrictive approach to regulating prosecutorial power with respect to that lever, the corresponding cell is shaded white; if it employs a broadly restrictive approach, the corresponding cell is shaded black; and if it employs one of the many intermediate regulatory approaches, the relevant cell is shaded gray. 220 More specifically, in the joinder–severance column, the black cell (Texas) represents nonjoinder, whereas white cells represent similar-offense joinder (the second-least-restrictive regulatory model). See supra section II.A; infra Appendix at Table A. In the preclusion column, the black cell (New Mexico) represents mandatory similar-offense joinder, whereas white cells represent the constitutional floor. See supra section II.B; infra Appendix at Table A. In the cumulative-sentencing column, the black cell (Texas, again) represents mandatory concurrent sentencing, whereas the remaining gray cells employ various discretionary regimes. See supra section II.C; infra Appendix at Table B. In the pretrial-evidentiary-review column, black cells represent states with hearings as of right, whereas the white cell (Indiana) represents a no-review state. See supra section III.A; infra Appendix at Table C. Finally, in the amendment column, the black cell (New York) represents a capped-amendment model, whereas the white cells represent unlimited amendment. See supra section IV.A; infra Appendix at Table E; see also supra note 193 (discussing California’s partial-ban regime, which is coded here as an intermediate regime); infra note 332 (same). Thus, examining the tables row by row, one can readily distinguish between those states that have “gaps” in their regulatory structures (that is to say, white cells), those that do not have any such gaps, and those that employ cutting edge regulatory models (that is to say, black cells) with respect to one or more procedural levers. As for the dependent variables, each table reports an outcome variable of particular interest: In Table 2, the states are organized in descending order according to the percentage of felony cases in which prosecutors are able to secure guilty pleas, while in Table 3 they are ranked according to the number of felony guilty pleas prosecutors obtain per every one hundred reported crimes. 221 The number of crimes per state is drawn from the FBI’s Uniform Crime Reports and reflects the combination of “violent crime” and “property crimes” reported in 2013. The plea, dismissal, and trial data for each state also come from 2013. For sources, see infra Appendix at Table F.

To be clear, these matrices simply collect some of the data that one would need to have on hand to begin examining the impact that specific procedural reforms might have on plea bargaining practices. Neither supports even preliminary correlational conclusions, given the lack of data concerning—among other things—the number of prosecutors employed in each jurisdiction, their budgetary constraints, the caseloads and budgetary constraints of defense attorneys, the types of crimes being prosecuted within each jurisdiction, and the magnitude of penalties authorized by each state’s substantive law. 222 Beyond these shortfalls, it is also important to note that the matrices do not offer a direct apples-to-apples comparison with respect to the dependent variables, as the plea rates are drawn from each state’s annual judicial performance reports, which do not use standardized units or frameworks for reporting their case-processing data. Perhaps most notably, diversionary agreements are not consistently reported in the same category across states; moreover, at least two states—Hawaii and Florida—report greater than ten percent of their felony cases as not being resolved by plea, trial, or dismissal. See infra Appendix at Table F (listing underlying sources and figures). Finally, as discussed in Parts II–IV, there is also considerable internal variability within even commonly coded procedural mechanisms, such that, for example, not all states with “indictment-bypass” regimes or “partially consecutive” sentencing regimes have truly comparable regulatory frameworks. See supra notes 112–113, 133–143 and accompanying text. And yet, the matrices suggest some of the promise of the research agenda that this Article invites. For one thing, they highlight a striking degree of variability with respect to both the independent and dependent variables: States differ not only in how they deploy individual procedural levers (as detailed in Parts II–IV above) but also with respect to their overarching regulatory approaches. Notably, some states (like Tennessee and Pennsylvania) employ more aggressive regulatory approaches across the board. At the same time, felony plea rates also vary considerably across states, with that variability driven by differences not in the rates of trials (which are fairly constant) but rather in the rates of dismissals—just as one would expect if procedural regulation is in fact driving resource-constrained prosecutors to screen their cases more aggressively. 223 The variability in plea rates reported here may be surprising to those readers familiar with the commonly repeated, but ultimately incorrect, assertion that “[r]oughly 95% of felony cases in the federal and state courts are resolved by guilty pleas.” Class v. United States, 138 S. Ct. 798, 807 (2018) (Alito, J., dissenting). The problem with this statement is that it fails to account for the significant number of cases resolved by dismissal. Thus, while it is true that approximately “[n]inety-seven percent of federal convictions and ninety-four percent of state convictions are the result of guilty pleas,” Lafler v. Cooper, 566 U.S. 156, 170 (2012) (emphasis added), only 89% of federal cases are resolved by guilty pleas, with the remainder resolved primarily by dismissals, as occurs in 8% of cases (as compared to the 3% of cases resolved by trial). See U.S. Courts, Federal Judicial Caseload Statistics 2017
Tables, tbl.D-4, [] (last visited Mar. 14, 2018). As for state courts, the data reported in Table 2 above show that plea bargaining rates and dismissal rates are quite variable, with plea rates ranging from 56% to 87% and dismissal rates ranging from 7% to 39%. See supra Table 2 (p. 1375). Of course, as noted in the text, omitting dismissals from any discussion of plea bargaining statistics substantially distorts the resulting account of how the criminal justice system is functioning: In a resource-constrained world, prosecutorial dismissals are a primary mechanism by which potentially beneficial case screening may take place, while judicial dismissals may be a sign of effective judicial oversight in action. See supra notes 60–61, 181 and accompanying text (discussing the relationship between procedural regulation, resource constraints, and prosecutorial screening of cases); supra Part III (discussing judicial screening and dismissals).
Moreover, while correlational inferences are undeniably premature, the charts suggest at least the facial plausibility of a causal relationship between plea bargaining’s subconsti­tu­tional procedural law and plea bargaining practices on the ground: Those states with the fewest “gaps” in their regulatory regimes and with the most aggressive regulatory models (that is, the fewest white cells and the most black cells) also seem to be those in which prosecutors have a comparatively harder time converting either cases or potential cases into pleas of guilt. 224 The four states without white-cell gaps are Illinois, Pennsylvania, Tennessee, and Vermont. Of these, three are in the bottom third of states when ranked by plea rates, with plea rates ranging from 55.7% (Tennessee) to 68.3% (Philadelphia, Pennsylvania). Similarly, the six states with one or more cutting-edge black cells are New Mexico, New York, Pennsylvania, Tennessee, Texas, and Wisconsin. And all six are in the bottom half of states when ranked by either plea rate or pleas per crime.

In sum, juxtaposing the legal frameworks uncovered here with data about criminal justice outcomes across jurisdictions surfaces important questions about how subconstitutional procedural law impacts the admin­istration of criminal justice. More importantly, it shows that answers to those questions are potentially within reach, particularly if scholars employ quantitative and qualitative analyses to study how these rules are incorporated into systemic plea bargaining practices on the ground. 225 For a powerful example of qualitative analysis in this domain, see Professor Mona Lynch’s recent comparative ethnography of plea bargaining practices in three different federal trial courts, Lynch, Hard Bargains, supra note 212. As for quantitative analyses, it bears noting that temporal variation may be particularly illuminating insofar as changes in plea rates (or the lack thereof) immediately surrounding changes to specific procedural levers could help isolate causal inferences. Cf. supra notes 55–56 (recounting Arkansas’s repeal of its nonjoinder rule in 2005, Texas’s adoption of such a rule in 1973, and Texas’s subsequent narrowing of that rule in 1997). Indeed, such a scholarly agenda holds the potential not only to enhance our understanding of a heretofore underappreciated regulatory framework but also to identify the rules within that framework that reformers should most “hope to change if we wish to change the outcomes produced by those rules.” 226 Kohler-Hausmann, supra note 212, at 267.

B. The Politics of Reform: Plea Bargaining’s Hidden Lawmakers

Plea bargaining’s hidden law, of course, did not write itself, which raises a final question that goes to the heart of plea bargaining’s underlying political economy: Who are the institutional actors—the lawmakers—most empowered to either facilitate or frustrate reforms in this arena, and what might be their motivations?

Here, too, this project offers an opportunity to revisit—and potentially to challenge—some of the dominant scholarly thinking about how our system of pleas is governed, including perhaps most notably the account so elegantly offered by the late Professor Bill Stuntz in The Pathological Politics of Criminal Law. 227 Stuntz, Politics, supra note 5. On Stuntz’s telling, prosecutors and legislators are driven by natural institutional incentives into a pathological alliance that leads to ever-increasing criminalization and incarceration. Prosecutors, Stuntz tells us, want to convict criminals in the most efficient manner possible. 228 Id. at 533–34. They therefore seek from legislators legal tools—such as expansive criminal codes—that will maximize their power and thereby make their jobs easier. 229 Id. at 537–38. And legislators, for their part, always oblige, because they know that they can be blamed only for withholding such tools—not for the prosecutors’ potential abuse of them down the road. 230 Id. at 537, 547–49. Hence we are left with an overbroad, overly punitive, and strikingly redundant substantive criminal law, which functions, according to Stuntz, as little more than a massive reallocation of power to prosecutors—to detrimental systemic effect. 231 Id. at 507–12; see also Barkow, Institutional Design, supra note 17, at 874–75; Barkow, Separation of Powers, supra note 11, at 1044–50.

One might fairly expect Stuntz’s analysis of plea bargaining’s political economy to carry over to the previously hidden legal frameworks excavated here. After all, if legislators are institutionally predisposed to enact substantive criminal laws that maximize prosecutorial power, why would they turn around and enact procedural constraints that cabin the very power they have just conferred? Conversely, to take the reformer’s perspective, if the political conditions for reform are ripe, why focus on the technocratic procedural mechanisms detailed here, rather than simply pressing for decriminalization and less punitive statutory sentences? 232 There may sometimes be tactical advantages to pursuing procedural reforms over substantive ones, particularly if procedurally grounded reforms are less salient and thus less likely to provoke a “soft on crime” backlash. Note, however, that while such an answer sounds in what Stuntz would call criminal law’s “surface politics,” which may well be changing, it is ultimately unresponsive to what Stuntz calls the “deep politics” of institutional incentives, summarized in the text. See Stuntz, Politics, supra note 5, at 510–11; cf. supra note 18 (discussing politics of reform).

This project reveals a surprising answer to these important questions: The institutions responsible for crafting these two bodies of law are often different. For while substantive criminal law is indeed authored by legisla­tures, the procedural law uncovered here is most frequently authored by judges—acting not in their familiar capacity as adjudicators deciding cases, but rather in a largely unacknowledged, quasi-legislative role, as drafters of their states’ Rules of Criminal Procedure. 233 Cf. Crespo, Systemic Facts, supra note 6, at 2069, 2115 (urging “a conceptualization of criminal courts that is broad enough” to see the ways in which they function as multifaceted “institutions of government,” tasked with administering core aspects of the American criminal justice system through practices—and through powers—beyond those that one might traditionally call “judicial”).

To appreciate the extent to which this unique and underappreciated lawmaking modality dominates plea bargaining’s hidden legal landscape, return for a moment to our very first subconstitutional lever: the law of joinder–severance. Only now notice not only how that law is substantively arrayed but also from whence it comes.

Figure 20: The Law of Joinder and Severance as a Law of Rules 234 For a detailed census, see infra Appendix at Table A (classifying each state’s joinder and severance provisions by lawmaking modality).


In this revised version of our familiar chart, each substantive silo is further subdivided to show the institutional source of the legal frameworks at issue. Thus, states that construct their joinder–severance regimes through legislative enactments are depicted in white, states with exclusively rule-based regimes are depicted in black, and the small handful of states with mixed frameworks are depicted in gray. 235 One state (South Carolina) constructs its joinder–severance regime entirely through its judicial case law and is thus not reflected in the figure. See id. Three additional states (Georgia, Kansas, and Oklahoma) are classified in the figure as statutory states even though portions of their substantive joinder–severance frameworks derive from judicial case law in the first instance. See id. In Table 4, infra (p. 1382–83), states that construct portions of their procedural regimes through judicial case law are reflected with an asterisk. As for those states classified as mixed, this category captures states in which rules of procedure and statutory provisions either combine or overlap to construct the governing joinder–severance regime. As the chart makes clear, court-made rules are a major component—indeed, the single largest compo­nent—of this body of law. And that pattern repeats itself with respect to most of the subconstitutional procedural levers outlined in this Article, as evidenced by the following tabular summary, which uses the same black and gray shading to depict procedural mechanisms authored wholly or partially by courts.

Table 4: Lawmaking Modality of Subconstitutional Procedural Law

STATE Piling On Overreaching Sliding Down
Joinder–Severance and Preclusion 236 For sources, see infra Appendix at Table A. Cumulative Sentencing 237 For sources, see infra Appendix at Table B. Pretrial Evidentiary Review 238 For sources, see infra Appendix at Table C. Amendment and Dismissal 239 For sources, see infra Appendix at Table E.
AK *
CA *
DE *
FL *
GA * *
HI *
ID *
IL * *
IN *
KS *
ME *
MD *
MA *
MI * *
MS *
MO *
NV *
NH * * *
NJ * *
NM * *
NC *
OH *†
OK *
OR *
SC * * * *
SD *
TX *
WA *
WV *
WI *
WY *

Exclusively Judicial (Black); Exclusively Legislative (White); Combination (Gray). Cells marked with an asterisk represent regimes established at least in part via judicial case law, while cells marked with a dagger represent regimes established at least in part via sentencing commissions, which are treated here as judicial bodies. 240 While state sentencing commissions can be difficult to classify, all of the commissions referenced in the Table are (with one exception) connected to their corresponding state judiciaries to at least some degree, whether by formal structure or by membership. The exception is the Oregon Criminal Justice Commission, which is separate from its state’s judiciary in both structure and membership and which is therefore not treated as a judicial body in the coding above. See Or. Rev. Stat. § 137.654 (2017). All other black and gray cells represent judicial lawmaking effectuated exclusively through rules of procedure.

Indeed, the power that state courts wield over plea bargaining’s hidden regulatory framework is even more dramatic than this visual depiction suggests. For courts are not only the primary source of plea bargaining’s quasi-legislative framework; in the majority of states, they are also the supreme legislative actor, empowered—quite unlike their federal counterparts—with authority to repeal or override legislatively enacted statutes simply by promulgating a countervailing rule of procedure. 241 In twenty-three states, judicial rulemaking authority is either exclusive or supreme in the procedural domain. For an accounting of the relevant state authorities, ordered alphabetically by state, see State ex rel. Collins v. Seidel, 691 P.2d 678, 682 (Ariz. 1984) (“[W]hen a conflict arises . . . we must draw the line. The legislature cannot repeal the Rules of . . . Procedure made pursuant to the [court’s constitutional] power.”); J.T. v. O’Rourke ex rel. Tenth Judicial Dist., 651 P.2d 407, 411 (Colo. 1982) (“[I]n cases of conflict, the court’s procedural rule would necessarily control [over] a procedural statute.”); In re Samantha C., 847 A.2d 883, 900 (Conn. 2004) (“[T]he general assembly lacks the power to enact rules governing procedure that is exclusively within the power of the courts.”); Del. Code tit. 11, § 5122 (2018) (“Any inconsistency or conflict between any rule of court . . . and any . . . statute of this State, dealing with practice and procedure in criminal actions in the Superior Court, shall be resolved in favor of such rule of court.”); State v. Currington, 700 P.2d 942, 944 (Idaho 1985) (“[W]here conflict exists between statutory criminal provisions and the Idaho Criminal Rules in matters of procedure, the rules will prevail.”); People v. Jackson, 371 N.E.2d 602, 603 (Ill. 1977) (invalidating a statute that “encroache[d] upon the rule-making power of th[e] court”); In re Pub. Law No. 305 & Pub. Law No. 309 of Indiana Acts of 1975, 334 N.E.2d 659, 662 (Ind. 1975) (“The procedural rules and the cases decided by the courts take precedence over any statute enacted concerning a procedural matter.”); State v. Mitchell, 672 P.2d 1, 9 (Kan. 1983) (“[W]hen court rules and a statute conflict . . . the court’s constitutional mandate [to enact rules of procedure] must prevail.”); Ky. Rev. Stat. Ann. § 447.154 (LexisNexis 2010) (“No . . . statute shall be construed . . . to limit in any manner the power of the Court of Justice to make rules governing practice and procedure in the courts.”); Eaton v. State, 302 A.2d 588, 592 (Me. 1973) (“[T]here can be no doubt of [the Supreme Judicial Court’s] power to override any procedural statutes that might be in conflict with [its] rules . . . .”); CAF Inv. Co. v. Saginaw Twp., 262 N.W.2d 863, 865 (Mich. 1972) (“Statutory rules of practice not in conflict with the court rules are effective until superseded by those rules.”); State v. Johnson, 514 N.W.2d 551, 553–54 (Minn. 1994) (“[T]he legislature ‘has no constitutional authority . . . to . . . enact statutes that will govern over court rules [of procedure] already in place.’” (alteration in original) (quoting Maynard E. Pirsig & Randall M. Tietjen, Court Procedure and the Separation of Powers in Minnesota, 15 Wm. Mitchell L. Rev. 141, 182 (1989))); Trull v. State, 811 So. 2d 243, 247 (Miss. Ct. App. 2000) (“[A]ny statute that conflicts with a rule established by the court is void.” (citing State v. Blenden, 748 So. 2d 77, 88–89 (Miss. 1999))); Lindauer v. Allen, 456 P.2d 851, 854 (Nev. 1969) (holding that when court rules and statutes conflict, the rule “must prevail”); State v. Leonardis, 375 A.2d 607, 614 (N.J. 1977) (interpreting the state constitution “to give the Court exclusive and plenary power over rules which are procedural in nature”); Sw. Cmty. Health Servs. v. Smith, 755 P.2d 40, 42 (N.M. 1988) (“[A]ny conflict between court rules and statutes that relate to procedure are today resolved by this Court in favor of the rules.”); N.D. Cent. Code § 27-02-09 (2016) (“[A]ll statutes relating to pleading, practice, and procedure in civil or criminal actions . . . have force and effect . . . unless and until amended or otherwise altered by rules promulgated by the supreme court.”); Ohio Const. art. IV, § 5(b) (“All laws in conflict with [court-issued] rules [of procedure] shall be of no further force or effect after such rules have taken effect.”); Pa. Const. art. V, § 10(c) (“All laws shall be suspended to the extent that they are inconsistent with [court-issued] rules [of procedure] . . . .”); 8 R.I. Gen. Laws § 8-6-2(a) (2012) (“[Court-issued] rules, when effective, shall supersede any statutory regulation in conflict therewith.”); Wash. Rev. Code § 2.04.200 (2017) (deeming “all laws in conflict” with “the rules of courts” to lack “force or effect”); State ex rel. Quelch v. Daugherty, 306 S.E.2d 233, 235 (W. Va. 1983) (“Legislative enactments which are not compatible with those prescribed by the judiciary or with its goals are unconstitutional violations of the separation of powers.”); White v. Fisher, 689 P.2d 102, 106 (Wyo. 1984) (“Matters dealing with procedure . . . are entirely within the province of this court[,] . . . [which has] constitutional authority . . . to issue rules of practice and procedure [that] . . . have stature in the hierarchy of law comparable to that of statutes enacted by the legislature.”).
In four additional states (Alaska, Arkansas, Florida, and Utah), judicial rulemaking authority supersedes ordinary statutes, but may be overruled by a two-thirds supermajority in the legislature. See Leege v. Martin, 379 P.2d 447, 450 (Alaska 1963) (“A two-thirds vote of . . . each house . . . is required . . . to change rules of practice and procedure[,] . . . a limitation [designed] to prevent . . . too easy intervention by the legislature . . . [into the] courts[’] . . . primary authority and responsibility for regulating [the courts’] own affairs.”); Ark. Const. amend. 80, § 9 (“Any rules promulgated by the Supreme Court . . . may be annulled or amended . . . by a two-thirds (2/3) vote of the membership of each house of the General Assembly.”); Johnson v. State, 336 So. 2d 93, 95 (Fla. 1976) (“[R]ules of practice and procedure . . . may be repealed by a general law enacted by a two-thirds vote of the Legislature, [but] the power to initiate them rests in this Court.”); Utah Const. art. VIII, § 4 (“The Legislature may amend the Rules of Procedure . . . adopted by the Supreme Court upon a vote of two-thirds of all members of both houses . . . .”).
Finally, in three additional states (Alabama, Maryland, and South Dakota), the legislature and the judiciary have co-equal lawmaking power, such that each can supersede or amend the other’s enactments. See Charles W. Gamble & Robert H. Fowlkes, Comprehensive Principles Governing the Interaction Between the Alabama Rules of Evidence and Alabama Statutory Rules of Evidence: The Need for Judicial Initiative and Legislative Deference When Both Branches of Government Hold Constitutional Power to Adopt Rules of Evidence, 56 Ala. L. Rev. 937, 956 (2005); Hauver v. Dorsey, 180 A.2d 475, 476 (Md. 1962) (“The Rules of this Court, of course, have the force of law, replacing inconsistent legislative provisions, subject only to the power of the Legislature to provide otherwise.”); Smith v. State, 248 A.2d 913, 914 n.1 (Md. Ct. Spec. App. 1969) (“[T]he rule prevails since it was adopted subsequent to the statute.”); In re Amendment of S.D.C.L. 23A-20-26, Rule 97-40 (S.D. Mar. 17, 1997), [] (amending a statute by court order).
For a now-somewhat-dated survey of state court rulemaking authority, see Donna J. Pugh, Judicial Rulemaking: A Compendium (1984).
That is to say, the familiar legal hierarchy is upended when it comes to state procedural law: Court-made rules trump legislature-enacted statutes more often than not. And as a result, the procedural frameworks that most directly regulate prosecutorial plea bargaining power reside not primarily in statutory codes, nor in the so-called “constitutional code of criminal procedure” drafted by the United States Supreme Court, but rather in an interlocking set of legal texts identical in form, substance, and function to legislation—save for the fact that they are enacted by state courts. 242 Cf. Henry J. Friendly, The Bill of Rights as a Code of Criminal Procedure, 53 Calif. L. Rev. 929, 953–54 (1965) (criticizing the Supreme Court for “applying the Bill of Rights” as if it “were a detailed code of criminal procedure”). Notably, even in states that do not afford their judiciaries legislative supremacy with respect to the promulgation of procedural rules, state courts can influence the procedural frameworks analyzed in this Article through their authority to interpret their respective state constitutions. See, e.g., William J. Brennan, Jr., State Constitutions and the Protection of Individual Rights, 90 Harv. L. Rev. 489, 500 (1977) (urging state courts to experiment with alternatives to federal Supreme Court precedent “even where the state and federal constitutions are similarly or identically phrased”); Goodwin Liu, State Constitutions and the Protection of Individual Rights: A Reappraisal, 92 N.Y.U. L. Rev. 1307, 1315 (2017) (“[S]tate courts, as the ultimate arbiters of state law, have the prerogative and duty to interpret their state constitutions independently.”). And yet, while such constitutional authority undoubtedly exists and is occasionally utilized to depart from the U.S. Supreme Court’s procedural baselines, state constitutional law is likely to be a more modest regulatory tool than the subconstitutional procedural law examined here, because much like federal constitutional law it operates under a perceived “countermajoritarian burden” that tends to “push[] toward judicial deference to political decisions” made by legislatures and executives. Paul W. Kahn, State Constitutionalism and the Problems of Fairness, 30 Val. U. L. Rev. 459, 471 (1996). By contrast, when state courts regulate through their rulemaking authority, they are not interpreting some preexisting constitutional text—they are authoring the governing law in the first instance. Subconstitutional procedural law thus offers courts an opportunity to craft regulatory frameworks in a manner that is both more straightforward and less fraught than is true of state constitutional law—a lawmaking modality that, perhaps for these reasons, tends to be used only sparingly. See James A. Gardner, The Failed Discourse of State Constitutionalism, 90 Mich. L. Rev. 761, 781 (1992) (recounting “a general unwillingness among state supreme courts to engage in any kind of analysis of the state constitution at all”).

And that, in turn, brings us back to Stuntz and the political economy of reform. For Stuntz’s tale of an unholy alliance between prosecutors and legislators is not without its imagined heroes. “[J]udges . . . alone,” he writes, “are likely to opt for narrower liability rules rather than broader ones,” for they “are much more likely than legislators or prosecutors to take the interests of defendants into account.” 243 Stuntz, Politics, supra note 5, at 510, 541; see also Barkow, Institutional Design, supra note 17, at 908 (“[G]reater judicial involvement would be the ideal corrective measure because it would interject a truly independent actor . . . less biased than a fellow prosecutor.”); Bibas, Outside the Shadow, supra note 17, at 2542 (urging reforms “to give judges a more active role in reviewing plea bargaining”). See generally Crespo, Systemic Facts, supra note 6, at 2060 (recounting courts’ “comparative institutional advantages . . . in the criminal justice arena”).
Plea bargaining scholars sometimes suggest that judges are unlikely to champion plea bargaining reform because (the argument goes) declining plea rates will put increased pressure on judicial resources. See, e.g., Barkow, Institutional Design, supra note 17, at 908 (“[I]t is precisely because of a heavy workload that judges have been complicit in the development of plea bargaining in the first place.”); Lynch, Administrative System, supra note 8, at 2142 (“The existing system . . . has arisen because the traditional adversarial model has become too expensive, contentious, and inefficient to be restored, at least given present . . . judicial resources.”); Wright & Miller, supra note 10, at 39–40 (“The judge, facing major caseload pressures, has little incentive to inquire behind the parties’ agreement. Indeed, sentencing judges tend to validate and encourage bargains . . . .”). This account of judicial incentives, however, is in some tension with the understanding of resource constraints advanced by Professors Bibas and Beirsbach, discussed supra notes 60–61. On that account, decreased plea rates are unlikely to produce more trials, for the simple reason that water cannot be had from a stone: Trial courts and prosecutors already stretched to capacity simply do not have the bandwidth to increase trial rates. Rather, to the extent that procedural reforms make plea bargains harder to obtain, any increased docket pressure is likely to be relieved through more prosecutorial dismissals and declinations—a potentially salutary outcome insofar as mass incarceration is concerned and an outcome that judges concerned solely with their own resource expenditures may well view agnostically. More fundamentally, the notion that judges will automatically resist procedural changes that increase the amount of time they spend presiding over trials rests on a potentially flawed assumption about how judges think—namely, that they would prefer to spend their days as a “rubber stamp” for “plea decisions,” Barkow, Institutional Design, supra note 17, at 872, rather than as the person in charge of a courtroom. See LaDoris H. Cordell, The Joy of Judging, 43 Harv. C.R.-C.L. L. Rev. 639, 643 (2008) (describing presiding over even simple trials as “the joy of judging”); cf. John R. Brown, In Memoriam: Judge J. Skelly Wright, 57 Geo. Wash. L. Rev. 1029, 1029 (1989) (“[T]he system demand[s] someone to be in charge, and what more appropriate party to take charge than the trial judge?”).
Indeed, it is his faith in judges that leads Stuntz to offer the prescription encountered throughout this Article: Criminal law, he says, should “be constitutionalized, with much more lawmaking power assigned to courts.” 244 Stuntz, Politics, supra note 5, at 579.

Judges on this telling are not merely heroes. They are heroes on the horizon—inclined to save the day, but denied the “constitutional warrant” to do so. 245 Id. at 588; see also id. at 510 (contending that courts are inclined, but unable, to “separate [the] natural allies,” that is, prosecutors and legislators); id. at 588 (“Any increase in judicial power over criminal law means an increase in constitutional power over criminal law.”); id. at 600 (“In order to have better criminal law, we need . . . some lawmaker other than legislatures or prosecutors. The most plausible lawmakers are the courts. The most plausible vehicle is the federal constitution.”). There is an irony, of course, in the fact that the Supreme Court—the authoring institution of the constitutional criminal procedure against which Stuntz writes—is itself composed of judges. But as a tribunal of only nine, its approach to criminal justice issues is far more subject to the historical contingency of its membership than can be said in the aggregate of the many thousands of lower state court judges. Cf. Crespo, Regaining Perspective, supra note 13, at 1990–95 (discussing potential biases of Supreme Court justices). By refocusing our view on subconstitutional law’s regulatory potential, however, this Article also reframes our understanding of the role that judges might play in plea bargaining’s political economy, for it reveals the extent to which state court judges already enjoy ample authority to regulate prosecutorial power. Indeed, they are frequently the primary and supreme enactors of the procedural codes through which such regulation takes place. And that realization, in turn, poses a test for Stuntz’s heroic account—because only one of two things can be true: Either the judges didn’t know that they already had this power, or they didn’t care.

The former possibility could well be true. After all, the nature of hidden law is that its significance can go unnoticed, even to those for whom it is most proximate. Indeed, the labor required to excavate these rules and to reconstitute them into a more clearly visible legal framework simply demonstrates how easily their regulatory upshot could go unnoticed, particularly if judges think of these rules primarily as governing what remains of our trial-based system of adjudication. 246 The rules, after all, are not called the Rules of Plea Bargaining Procedure. Cf. Stephanos Bibas, Designing Plea Bargaining from the Ground Up: Accuracy and Fairness Without Trials as Backstops, 57 Wm. & Mary L. Rev. 1055, 1059 (2016) (“This plea-bargaining system was not planned, but jury-rigged. It grew up below the radar . . . . [Accordingly, the few legal] safeguards that exist [within it] are largely designed to ensure the centrality of jury trials.”); Ortman, supra note 1, at 516 (suggesting that “perhaps policymakers never reexamined the fit between charging standards and the plea bargaining regime,” such “that probable cause’s survival is nothing more (or less) than path dependency”); Roberts, supra note 172, at 330 (suggesting that equitable dismissal rules may be unfamiliar to academics and practitioners alike). Alternatively, judges might not be aware of how differently these rules are structured across jurisdictions—and thus may not appreciate the full scope of their regulatory potential. But of course, if judges have by and large been unaware of the significance these provisions hold, then any efforts to focus attention on these subconstitutional proce­dural levers should have the salutary effect of putting Stuntz’s heroic vision to the test, prompting judges to consider corrective reforms that he mistakenly thought only constitutional authority could afford.

Or the alternative possibility could hold: Perhaps these levers will be brought to courts’ attention and nothing will change. In that event, one might wonder whether judges, unlike academics, were aware of these regulatory tools all along—and simply not using them. Such a response would suggest that scholarly skeptics of Stuntz’s heroic judicial account may have the better angle on the problem: Far from being the heroes on the horizon, our judges may be complicit enablers of plea bargaining’s current pathology. 247 For examples of such skeptical accounts, see Benjamin Levin, Values and Assumptions in Criminal Adjudication, 129 Harv. L. Rev. Forum 379, 383 (2016),
wp-content/uploads/2016/06/vol129_B-Levin.pdf [] (arguing that “normative preferences or priors [are] to blame for courts’ shortcomings”); id. at 385 n.35 (citing Gregory L. Acquaviva & John D. Castiglione, Judicial Diversity on State Supreme Courts, 39 Seton Hall L. Rev. 1203, 1235 tbl.10 (2009), for the proposition “that judges with prosecutorial experience tend to outnumber those with criminal defense experience”); see also Meares, supra note 17, at 862 n.36 (noting federal judges “rarely will exercise the power” they enjoy under existing procedural rules to oversee charge-sliding amendments); cf. Crespo, Systemic Facts, supra note 6, at 2112–14 (observing that procedural tools designed to “facilitate systemic judicial review” “cannot replace the exercise of careful judgment when it comes to the core normative questions at the heart of the criminal justice system” or “force judges to abandon any preexisting views they might hold”). If the core problem is in fact lack of judicial will to improve systemic pathologies, the solution may be “to change the manner in which such judges are selected, or the judges selected through that process.” Crespo, Systemic Facts, supra note 6, at 2114.

Ultimately then, this Article might be seen not only as calling for further study of procedural law’s potential practical effects but also as something of an experiment in its own right: By putting to rest the illusion, perhaps shared by judges themselves, that judges lack the tools to regulate prosecutorial power, it creates an opportunity to observe whether and how judges put these newly uncovered tools to use. Time, as they say, will tell.


Our criminal justice system is a system of pleas. Few who know it well think that it is working. For decades, some of the leading scholarly voices pressing for reform have operated under the assumption that plea bargain­ing operates “beyond the shadow of the law,” given the minimal role that substantive and constitutional criminal law play in regulating prosecutorial power. But this Article has shown that assumption to be misguided: Prosecutorial plea bargaining power may operate beyond the shadow of the law we know well, but it is comprehensively and inextricably intertwined with a separate, heretofore hidden body of law that not only regulates numerous aspects of plea bargaining’s dynamics but in fact establishes the mechanisms of such power in the first instance. Far from some abstract or theoretical set of legal constructs, these procedural regimes exist in various different forms across our pluralist criminal justice system, which is peppered with examples highlighting the extent to which law reform is both possible and, in some instances, already underway. Whether reform centered on these newly excavated legal levers will actually help ameliorate plea bargaining’s current pathologies is an open question, the answer to which is dependent in no small part on the willingness and ability of scholars, reformists, and policymakers alike to see what is before them—a hidden law of plea bargaining, awaiting their attention.


**I am grateful to the many diligent research assistants who helped compile and analyze these sources: Paulina Arnold, Jeffrey Campbell, Colin Doyle, Henry Druschel, Mary Goetz, Max Gottschall, Victoria Hall-Palerm, Andrew Leon Hanna, Sarah Kahwash, Benjamin Lewis, Paul Maneri, Joshua Olszewski-Jubelirer, Isaac Park, Bradley Pough, Charlotte Robinson, Jacque Sahlberg, Matthew Scarvie, William Schmidt, Imani Tisdale, Emily Villa, and Olivia Warren.

The tables that follow capture the subconstitutional (and in some instances the constitutional) procedural law of the states with respect to the various procedural levers discussed in this Article. 248 Cf. supra note 242 (discussing state constitutional law). The applicable law of the federal system is also reported at the end of each table for com­par­i­son. 249 Cf. supra note 21 (discussing federal law). Together, the surveys offer a grounded account of the procedural heterogeneity employed across the country—a snapshot of plea bargaining’s hidden regulatory levers as they stood at the time the surveys generating this information were conducted. The tables do not, however, offer a comprehensive or fully nuanced almanac of the hundreds of procedural doctrines at issue. State law on these topics is complex, variable, and susceptible to change over time. Readers seeking more detailed portraits of any given state should therefore treat the wide-angle account offered here as an initial point of entry to further inquiry, not a final stop.

Throughout the tables that follow, legal authorities constituting the corresponding procedural law within each state are reported first, with the authorities organized into columns based on their lawmaking modality (that is to say, by whether they are court-made rules, legislature-enacted statutes, etc.). Each state’s procedural lever is then assigned a substantive code that taxonomizes the procedural landscape at issue. An overview of these codes is provided in an introductory note accompanying each table.

As a general rule, the surveys that produced these tables focused on the sources of authority that originated the procedural rules at issue; supplemental case law interpreting such provisions was examined for clarity as needed and is occasionally discussed in the notes that follow, but the survey of interpretive case law within each state was not exhaustive.

Table A: The Law of Joinder, Severance, and Preclusion

This table supports Figures 4 and 8, supra (pp. 1321, 1331). The joinder column reports the outer limit of joinable offenses for each state. The severance column reports whether that joinder authority is constrained solely by an equitable failsafe or whether the state instead grants defendants a form of severance as of right (AOR). When severance as of right is available, the severance column reports the narrowest relationship among joined charges for which that right exists. Thus, “Similar Offense AOR” refers to a regime in which the defendant enjoys a right to sever any charges that are joined only by virtue of the fact that they are similar offenses or that lack even that nexus to one another; a defendant in such a regime has no right to sever offenses joined by virtue of their enjoying a closer factual nexus (for example, by virtue of their being part of the same event, the same chain of events, or a common scheme). The composite joinder–severance column describes the prosecutor’s joinder authority as constrained by the defendant’s right to sever if such a right exists—that is to say, it reports the scope of joinder authority that is not subject to a defendant’s severance veto. This column will match the joinder column in states with only equitable-failsafe severance; states in which severance as of right constrains joinder are marked in the composite column with an asterisk for ease of reference. Turning to preclusion, the penultimate column reports the scope of the “preclusive shadow” associated with an initial prosecution in the state. When the scope of preclusion matches the baseline provided by the federal Constitution’s Double Jeopardy Clause, the state is coded as adhering to the “constitutional floor,” which it may do either by expressly embracing the federal doctrine as a matter of state constitutional or subconstitutional law, or by leaving federal constitutional law to operate as the exclusive constraint. Finally, the last column describes the relationship between the state’s law of preclusion and its corresponding joinder–severance framework, a relationship that yields either mandatory- or permissive-joinder regimes. Note that, throughout this table, the line between “Same Event” and “Event Chain” models navigates ambiguous terrain, as states take varying and often hazy approaches to defining what constitutes a factual “event.” The decision to assign one of these codes over the other may thus at times suggest a degree of clarity not actually present in the underlying law of the jurisdiction.

State Source of Authority Procedural Regime
Court Rule Statute Case Law Joinder Severance Joinder–Severance Composite Scope of Preclusion Mandatory vs. Permissive Joinder
AL Ala. R. Crim. P. 13.3 (joinder); id. 13.4 (severance) Ala. Code § 15-8-52 (LexisNexis 2015) (joinder) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
AK Alaska R. Crim. P. 8 (joinder); id. 14 (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
AZ Ariz. R. Crim. P. 13.3 (joinder); id. 13.4 (severance) Similar Offense Similar Offense AOR Common Scheme* Constitutional Floor Permissive
AR Ark. R. Crim. P. 21.1 (joinder); id. 22.2 (severance); id. 21.3 (preclusion) Similar Offense Similar Offense AOR Common Scheme* Event Chain Permissive
CA Cal. Penal Code § 954 (2018) (joinder–severance) Kellett v. Superior Court of Sacramento Cty., 409 P.2d 206, 210 (Cal. 1966) (preclusion) Similar Offense Equitable Failsafe Similar Offense Event Chain Permissive
CO Colo. R. Crim. P. 8(a)(2) (joinder); id. 14 (severance); id. 8(a) (preclusion) Colo. Rev. Stat. § 18-1-408(4) (2017) (severance); id. § 18-1-408(2) (preclusion) Similar Offense Equitable Failsafe Similar Offense Event Chain Permissive
CT Conn. R. Super. Ct. § 36-21 (joinder); id. § 41-18 (severance) Conn. Gen. Stat. Ann. § 54-57 (West 2009) (joinder) Unlimited Equitable Failsafe Unlimited/
Similar Offense 250 The text of the cited statute limits joinder to “offenses of the same character.” Conn. Gen. Stat. Ann. § 54-57. The text of the first cited rule, however, establishes an unlimited-joinder regime: “Two or more offenses may be charged in the same information in a separate count for each offense for any defendant.” Conn. R. Super. Ct. § 36-21. According to the state supreme court, language like this that “omits reference to the requirement that the offenses joined be of the ‘same character’” creates an “extremely broad provision” under which determining “whether the offenses are of the ‘same character’ is no longer essential.” State v. King, 445 A.2d 901, 904 (Conn. 1982). Moreover, the court has made clear that if the rule and the statute conflict, the rule prevails. See id. (“[T]he general assembly lacks the power to enact rules governing procedure . . . . Therefore, we will measure the trial court’s decision on the motion for joinder against the [rule].”). For these reasons, Connecticut is classified as an unlimited-joinder regime in Figure 4, supra (p. 1321). It bears noting, however, that case law in the jurisdiction appears to have focused primarily on the relationship between the cited joinder statute and a separate joinder rule, which “governs the joinder of multiple informations for trial,” State v. Payne, 34 A.3d 370, 379 (Conn. 2012) (citing Conn. R. Super. Ct. § 41-19), rather than the rule cited here, which pertains to the number of charges that “may be charged in the same information,” Conn. R. Super. Ct. § 36-21. Indeed, in an opinion that does not mention Rule 36-21 at all, the state supreme court has held that the joinder statute “governs the circumstances under which [prosecutors] may join multiple charges in a single information,” and thus limits such power to offenses “of the ‘same character.’” Payne, 34 A.3d at 379–80 (quoting Conn. Gen. Stat. Ann. § 54-57). Suffice to say, however one might reconcile Connecticut’s various joinder provisions, the state embraces either an “extremely broad” unlimited-joinder regime, King, 445 A.2d at 904, or a slightly narrower similar-offense regime.
Constitutional Floor Permissive
DE Del. Super. Ct. R. Crim. P. 8 (joinder); id. 14 (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
FL Fla. R. Crim. P. 3.150(a) (joinder); id. 3.152 (severance); id. 3.151(c) (preclusion) Event Chain Equitable Failsafe Event Chain Event Chain Mandatory
GA Ga. Code Ann. § 16-1-7 (2011) (joinder–severance, preclusion) Dingler v. State, 211 S.E.2d 752, 753 (Ga. 1975) (joinder) Similar Offense Similar Offense AOR Common Scheme* Event Chain 251 See State v. Pruiett, 751 S.E.2d 579, 582 (Ga. Ct. App. 2013) (interpreting “same conduct” in the preclusion statute as synonymous with “same transaction”). Permissive
HI Haw. R. Penal P. 8(a) (joinder); id. 14 (severance) Haw. Rev. Stat. Ann. § 806-22 (LexisNexis 2016) (joinder); id. § 701-109(3) (severance) Similar Offense Equitable Failsafe Similar Offense Same Event 252 See State v. Keliiheleua, 95 P.3d 605, 612 (Haw. 2004) (interpreting the statutory phrase “same episode” to refer to charges “‘so closely related in time, place and circumstances that a complete account of one charge cannot be related without referring to details of the other charge’” (quoting State v. Servantes, 804 P.2d 1347, 1349 (Haw. 1991))). Permissive
ID Idaho Crim. R. 8(a) (joinder); id. 14 (severance) Idaho Code § 19-1432 (2017) (joinder)


Common Scheme 253 The statute allows for more permissive joinder than the rule, but State v. Currington indicates that when a rule and statute conflict on a procedural matter, the rule prevails. 700 P.2d 942, 944 (Idaho 1985). Equitable Failsafe Common Scheme Constitutional Floor Permissive
IL 725 Ill. Comp. Stat. Ann. 5/111-4 (West 2006) (joinder); id. 5/114-8 (severance); 720 Ill. Comp. Stat. Ann. 5/3-3 (West 2016) (preclusion) Event Chain Equitable Failsafe Event Chain Same Event Permissive
IN Ind. Code § 35-34-1-9 (2017) (joinder); id. § 35-34-1-11 (severance); id. § 35-34-1-10 (preclusion); id. § 35-41-4-4 (preclusion) Similar Offense Similar Offense AOR Common Scheme* Common Scheme 254 The preclusion statute bars successive prosecutions “for an offense with which the defendant should have been charged in the former prosecution.” Ind. Code § 35-41-4-4. “The Indiana Supreme Court has held that the words ‘should have been charged’ . . . must be read in conjunction with Indiana’s joinder statute . . . .” Hahn v. State, 67 N.E.3d 1071, 1082 (Ind. Ct. App. 2016). Accordingly, a prosecutor who attempts “to bring multiple prosecutions for a series of acts” that “are part of a single scheme or plan” does so “at [her] own peril,” as she may be barred from pursuing any second or subsequent prosecutions by the preclusion statute. Id. (internal quotation marks omitted) (quoting Williams v. State, 762 N.E.2d 1216, 1219 (Ind. 2002)); see also id. at 1082–83 (“To determine whether contemporaneous crimes are part of a single scheme or plan, we examine ‘whether they are connected by a distinctive nature, have a common modus operandi, and a common motive.’” (internal quotation marks omitted) (quoting Williams, 762 N.E.2d at 1220)). Mandatory
IA Iowa Ct. R. 2.6 (joinder–severance) Common Scheme Equitable Failsafe Common Scheme Constitutional Floor Permissive
KS Kan. Stat. Ann. § 22-3202 (2007) (joinder) State v. Shaffer, 624 P.2d 440, 443 (Kan. 1981) (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
KY Ky. R. Crim. P. 6.18 (joinder); id. 8.31 (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
LA La. Code Crim. Proc. Ann. art. 493 (2017) (joinder);
id. art. 495.1 (severance)
Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
ME Me. R. Unified Crim. P. 8 (joinder–severance) Me. Rev. Stat. Ann. tit. 17-A, § 14 (2006) (preclusion) Similar Offense Equitable Failsafe Similar Offense Event Chain Permissive
MD Md. R. 4-203 (joinder); id. 4-253 (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
MA Mass. R. Crim. P. 9 (joinder–severance) Common Scheme Equitable Failsafe Common Scheme Constitutional Floor Permissive
MI Mich. Ct. R. 6.120(a) (joinder); id. 6.120(b)–(c) (severance) Unlimited Similar Offense AOR Common Scheme* Constitutional Floor Permissive
MN Minn. R. Crim. P. 17.03 (joinder–severance) Minn. Stat. § 609.035 (2017) (preclusion) Unlimited Common Scheme AOR 255 See State v. Profit, 591 N.W.2d 451, 458–60 (Minn. 1999) (interpreting the right to sever “charges [that] are not related” to apply when the offenses are “not part of a single behavioral incident or course of conduct”). Event Chain* Same Event 256 See State v. Bakken, 883 N.W.2d 264, 270 (Minn. 2016) (interpreting statutory-preclusion language to apply to offenses arising from “a single behavioral incident,” defined by “whether ‘the offenses occurred at substantially the same time and place,’ and . . . ‘[were] motivated by an effort to obtain a single criminal objective’” (first quoting State v. Jones, 848 N.W.2d 528, 533 (Minn. 2014); then quoting State v. Bauer, 792 N.W.2d 825, 828 (Minn. 2011))). Permissive
MS Miss. R. Crim. P. 14.2 (joinder); id. 14.3 (severance) Miss. Code Ann. § 99-7-2 (2015) (joinder) Common Scheme Equitable Failsafe Common Scheme Constitutional Floor Permissive
MO Mo. Sup. Ct. R. 23.05 (joinder); id. 24.07 (severance) Mo. Rev. Stat. § 545.140 (2017) (joinder) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
MT Mont. Code Ann. § 46-11-404 (2017) (joinder); id. § 46-13-211 (severance); id. § 46-11-503 (preclusion) Similar Offense Equitable Failsafe Similar Offense Common Scheme 257 See State v. Hocevar, 7 P.3d 329, 351–52 (Mont. 2000) (interpreting statutory preclusion of offenses arising from the “same transaction” to apply broadly to any offenses sharing a common “criminal objective” (citing Mont. Code Ann. § 46–1–202(23)). Permissive
NE Neb. Rev. Stat. § 29-2002 (2016) (joinder–severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
NV Nev. Rev. Stat. § 173.115 (2015) (joinder); id. § 174.165 (severance) Common Scheme Equitable Failsafe Common Scheme Constitutional Floor Permissive
NH N.H. R. Crim. P. 20 (joinder–severance, preclusion) Common Scheme 258 New Hampshire’s joinder rule directly incorporates an equitable failsafe more typically provided by the law of severance. See N.H. R. Crim. P. 20(a)(2) (permitting prosecutors to “move for joinder” while providing that the “judge shall join the charges . . . unless [she] determines that joinder is not in the best interests of justice”). As to the scope of the joinder rule, the state permits offenses to be charged together when they are part of the same “criminal episode,” part of a “common scheme,” or are otherwise “logically and factually connected in a manner that does not solely demonstrate that the accused has a propensity to engage in criminal conduct.” Id. 20(a)(1)(C). Equitable Failsafe Common Scheme Event Chain Permissive
NJ N.J. Ct. R. 3:7-6 (joinder); id. 3:15-2 (severance); id. 3:15-1 (preclusion) N.J. Stat. Ann. § 2C:1-8 (West 2015) (preclusion) Similar Offense Equitable Failsafe Similar Offense Common Scheme 259 See State v. Williams, 799 A.2d 470, 476 (N.J. 2002) (interpreting statutory preclusion of charges arising from the “same episode” to apply “even when the defendant’s actions occurred over several months and at different locations, [if] the events were connected as part of a larger scheme”). Permissive
NM N.M. R. 5-203; 6-306; 7-306 (joinder–severance); id. 5-203 (preclusion) Similar Offense Equitable Failsafe Similar Offense Similar Offense Mandatory
NY N.Y. Crim. Proc. Law § 200.20 (McKinney 2007) (joinder–severance); N.Y. Crim. Proc. Law § 40.40 (McKinney 2016) (preclusion) Similar Offense Equitable Failsafe 260 New York’s equitable failsafe exists only when charges are joined as similar offenses. Similar Offense Common Scheme 261 Cf. N.Y. Crim. Proc. Law § 40.20(2) (enumerating exceptions to the preclusion rule). Permissive
NC N.C. Gen. Stat. § 15A-926 (2017) (joinder, preclusion); id. § 15A-927 (severance) Common Scheme Equitable Failsafe Common Scheme Common Scheme Mandatory
ND N.D. R. Crim. P. 8 (joinder); id. 14 (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
OH Ohio R. Crim. P. 8 (joinder); id. 14 (severance) Ohio Rev. Code Ann. § 2941.04 (West 2006) (joinder–severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
OK Okla. Stat. Ann. tit. 22, § 439 (West 2003) (severance) Glass v. State, 701 P.2d 765, 768 (Okla. 1985) (joinder) Event Chain Equitable Failsafe Event Chain Constitutional Floor Permissive
OR Or. Rev. Stat. § 132.560 (2015) (joinder–severance); id. § 131.505 (preclusion); id. § 131.515 (preclusion) Similar Offense Equitable Failsafe Similar Offense Same Event Permissive
PA Pa. R. Crim. P. 563, 582 (joinder); id. 583 (severance) 18 Pa. Stat. and Cons. Stat. Ann. § 110 (West 2015) (preclusion) Common Scheme 262 See Commonwealth v. Larkins, 449 A.2d 42, 44–45 (Pa. 1982) (interpreting Pennsylvania law to permit joinder of offenses occurring over a three-week period because each showed evidence of a common scheme). Equitable Failsafe Common Scheme Event Chain Permissive
RI R.I. Super. Ct. R. Crim. P. 8(a) (joinder); id. 14 (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
SC State v. Simmons, 573 S.E.2d 856, 860–61 (S.C. Ct. App. 2002) (joinder–severance) Event Chain Equitable Failsafe Event Chain Constitutional Floor Permissive
SD S.D. Codified Laws § 23A-6-23 (2016) (joinder); id. § 23A-11-2 (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
TN Tenn. R. Crim. P. 8 (joinder, preclusion); id. 14 (severance) Similar Offense Similar Offense AOR Common Scheme* Same Event 263 See State v. Johnson, 342 S.W.3d 468, 475 (Tenn. 2011) (interpreting the rule precluding offenses arising from “same criminal episode” as a prior prosecution as covering offenses that “occur simultaneously or in close sequence and . . . in the same place or in closely situated places” such that “proof of one offense necessarily involves proof of the others”). Permissive
TX Tex. Penal Code § 3.02 (2017) (joinder); id. § 3.04 (severance); Tex. Code Crim. Proc. art. 21.24 (2017) (joinder) Similar Offense 264 While Tex. Penal Code § 3.02 limits joinder to charges arising out of the same “criminal episode,” Tex. Penal Code § 3.01 defines “criminal episode” as “the commission of two or more offenses . . . [1] pursuant to the same transaction or pursuant to two or more transactions that are connected or constitute a common scheme or plan; or . . . [2] the repeated commission of the same or similar offenses.” Universal AOR 265 Cf. Tex. Penal Code § 3.03(b) (enumerating exceptions to the defendant’s right to sever); supra note 55 (discussing exceptions to Texas’s severance rule and its requirement that a defendant choose between whether to invoke the state’s severance rule or its concurrent-sentencing rule). Nonjoinder* Constitutional Floor Permissive
UT Utah Code Ann. § 77-8a-1 (LexisNexis 2012) (joinder–severance); id. § 76-1-402 to -403 (preclusion) Common Scheme Equitable Failsafe Common Scheme Same Event Permissive
VT Vt. R. Crim. P. 8 (joinder); id. 14 (severance); id. 13 (preclusion) Similar Offense Similar Offense AOR Common Scheme* Event Chain Permissive
VA Va. Sup. Ct. R. 3A:6 (joinder); id. 3A:10 (severance) Common Scheme Equitable Failsafe Common Scheme Constitutional Floor Permissive
WA Wash. Super. Ct. Crim. R. 4.3 (joinder); id. 4.4 (severance); id. 4.3.1 (preclusion) Similar Offense Equitable Failsafe Similar Offense Event Chain 266 See State v. Lee, 939 P.2d 1223, 1226 (Wash. 1997) (en banc) (interpreting rule language precluding offenses arising from the “‘same conduct’” as covering “offenses based upon the . . . same series of physical acts,” even if those acts “span a period of time and involve more than one place” (quoting Wash. Super. Ct. Crim. R. 4.3)). Permissive
WV W. Va. R. Crim. P. 8 (joinder); id. 14 (severance); id. 8 (preclusion) Similar Offense Equitable Failsafe Similar Offense Common Scheme Permissive
WI Wis. Stat. § 971.12 (2018) (joinder–severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
WY Wyo. R. Crim. P. 8 (joinder); id. 14 (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive
FED Fed. R. Crim. P. 8(a) (joinder); id. 14 (severance) Similar Offense Equitable Failsafe Similar Offense Constitutional Floor Permissive


Table B: The Law of Cumulative Sentencing

This table supports Figures 10 and 11, supra (pp. 1336, 1337). The “primary sentencing regime” column describes the generic framework governing cumulative sentencing within the state—that is to say, the sentencing regime that operates in the absence of specific provisions creating carve-outs to that generic approach. For a description of the substantive codes used to classify these generic frameworks, see supra notes 90–94 and accompanying text. Most states employ carve-out exceptions to their generic sentencing regimes. Accordingly, the “subsidiary sentencing regime” column reports whether carve-outs within the state impose mandatory consecutive sentences, mandatory concurrent sentences, or a combination of mandatory consecutive sentences in some instances and mandatory concurrent sentences in others (coded as “both”). In the rare instance in which a state does not employ carve-outs to its general regime, the cell in the subsidiary column is coded “n/a.” The subconstitutional law of merger, discussed supra note 97, is not surveyed here.

State Source of Authority Procedural Regime
Court Rule Statute Case Law Sentencing Guidelines Primary Sentencing Regime Subsidiary Sentencing Regime
AL Ala. R. Crim. P. 26.12 Ala. Code § 14-4-9 (LexisNexis 2015) Presumptive and Voluntary Sentencing Standards Manual 26 (Ala. Sentencing Comm’n 2016) 267 Presumptive and Voluntary Sentencing Standards Manual 26 (Ala. Sentencing Comm’n 2016),


Open-Ended Discretion
(Default Consecutive)
Partially Mandatory Concurrent 268 While the cited rule and statutory provisions create a default-consecutive sentencing regime, the Alabama sentencing guidelines impose a mandatory-concurrent regime. See id. at 26 (“The total or aggregate sentence for the sentencing event cannot exceed the sentence selected for the most serious offense.”); id. at 23 (“A sentencing event includes all convictions sentenced at the same time, whether included as counts in one case or in multiple cases . . . .”). The guidelines, however, are “presumptive sentencing recommendations for non-violent offenses and voluntary sentencing recommendations for violent offenses,” id. at 16, and thus formally constrain the discretion otherwise granted by rule and statute in only a subset of cases. Cf. Ala. Code § 12-25-34.2(b) (LexisNexis 2012) (requiring adherence to presumptive sentences in the absence of aggravating or mitigating factors); Presumptive and Voluntary Sentencing Standards Manual, supra note 267, at 29–30.
AK Alaska Stat. § 12.55.127 (2016) Open-Ended Discretion Partially Mandatory Consecutive
AZ Ariz. R. Crim. P. 26.13 Ariz. Rev. Stat. Ann. §§ 13-708, 13-711, 13-1212, 13-1304, 13-2503 (2010) Open-Ended Discretion
(Default Consecutive)
Partially Mandatory Consecutive
AR Ark. Code Ann. § 5-4-403 (2013) Open-Ended Discretion
(Default Concurrent)
Partially Mandatory Concurrent
CA Cal. R. Ct. 4.406, 4.425 Cal. Penal Code §§ 667.6(d), 669, 1170.1, 1170.12, 1202.2 (2018) Structured Discretion Both 269 California imposes a partial cap on consecutive sentences that effectively mandates partially concurrent sentences when the cap applies. See Cal. Penal Code § 1170.1 (limiting length of “subordinate” consecutive terms); id. §1170.12(a)(1) (stating exception).
CO Colo. Rev. Stat. §§ 18-1-408, 18-1.3-406, 18-3-402(5), 18-8-209, 18-8-212, 18-12-108 (2017) Mandatory Concurrent Partially Mandatory Consecutive
CT Conn. Gen. Stat. Ann. §§ 53-396, 53a-37, 53a-59b (West 2012) Open-Ended Discretion Partially Mandatory Consecutive
DE Del. Code Ann. tit. 11, § 3901 (2018) Open-Ended Discretion 270 From 1976 through 2014, Delaware had a mandatory-consecutive regime. See Fountain v. State, 139 A.3d 837, 840 (Del. 2016) (“In 1976, the Delaware legislature eliminated concurrent sentencing entirely . . . .”); id. at 840–41 (discussing the 2014 statute that “restore[d] judicial discretion to permit the imposition of either concurrent or consecutive sentences” (emphasis omitted) (internal quotation marks omitted) (quoting H.B. 312 synopsis, 147th Gen. Assemb., Reg. Sess. (Del. 2014),
BillDetail?LegislationId=22829 [])).
Partially Mandatory Consecutive
FL Fla. R. Crim. P. 3.701(d)(2)–(3) Fla. Stat. §§ 784.048, 921.16, 944.40. 944.48 (2017) Open-Ended Discretion
(Default Concurrent) 271 Florida employs a default-concurrent regime for offenses charged in a single case and a default-consecutive regime for offenses charged in separate cases. See Fla. Stat. § 921.16.
Partially Mandatory Consecutive
GA Ga. Code Ann. § 17-10-10 (2013); Ga. Code Ann. § 16-11-106 (2011) Open-Ended Discretion
(Default Concurrent)
Partially Mandatory Consecutive
HI Haw. Rev. Stat. Ann. §§ 706-606, 706-668.5, 712-1240.5 (LexisNexis 2016) Structured Discretion Partially Mandatory Consecutive
ID Idaho Code §§ 18-308, 18-915, 18-2505(1) (2017) Open-Ended Discretion Partially Mandatory Consecutive
IL 730 Ill. Comp. Stat. Ann. 5/5-8-4 (West 2007) Open-Ended Discretion
(Default Concurrent) 272 The generic sentencing regime requires that “sentences shall run concurrently,” unless the sentencing judge determines “that consecutive sentences are required to protect the public” due to “the nature and circumstances of the offense and the history and character of the defendant.” 730 Ill. Comp. Stat. Ann. 5/5-8-4(a)–(c).
IN Ind. Code § 35-50-1-2(c)–(f) (2017) Owens v. State, 916 N.E.2d 913, 917 (Ind. Ct. App. 2009) Structured Discretion 273 See Sanquenetti v. State, 727 N.E.2d 437, 442 (Ind. 2000) (“Aggravating circumstances may include, but are not limited to, any of several statutorily enumerated factors.” (citing Ind. Code § 35-38-1-7.1)). Both
IA Iowa Ct. R. 2.23 Iowa Code §§ 901.8, 908.10 (2018). Open-Ended Discretion Partially Mandatory Consecutive
KS Kan. Stat. Ann. §§ 21-6606, 21-6819 (2016). Open-Ended Discretion
(Default Concurrent)
KY Ky. R. Crim. P. 11.04 Ky. Rev. Stat. Ann. §§ 532.110, 533.060 (LexisNexis 2014) Open-Ended Discretion
(Default Concurrent)
LA La. Code Crim. Proc. Ann. art. 883 (2008) Open-Ended Discretion
(Default Concurrent) 274 Louisiana employs a default-concurrent regime for offenses based on the same transaction and a default-consecutive regime for offenses based on different transactions.
ME Me. R. Unified Crim. P. 35(d) Me. Rev. Stat. Ann. tit. 17-A, § 1256 (2006) Structured Discretion Both
MD Md. R. 4-351 Md. Code Ann., Corr. Servs. §§ 7-403, 9-201 (LexisNexis 2017); Md. Code Ann., Crim. Law §§ 4-204, 4-306, 5-608.1, 9-407 (LexisNexis 2017) Md. Sentencing Guidelines Manual Ch. 9 (Md. State Comm’n on Criminal Sentencing Policy 2017) 275 Md. Sentencing Guidelines Manual Ch. 9 (Md. State Comm’n on Criminal
Sentencing Policy 2017), [].
Open-Ended Discretion
(Default Concurrent) 276 While the Maryland Sentencing Guidelines impose a mandatory-concurrent regime, the guidelines themselves “are voluntary sentencing guidelines that a court need not follow.” Md. Code Ann., Crim. Proc. § 6 -211 (LexisNexis 2016). Accordingly, this state has been coded as a “default-concurrent” regime.
Partially Mandatory Consecutive
MA Mass. Ann. Laws ch. 90, § 24V (LexisNexis 2012) Advisory Sentencing Guidelines 44–47 (Mass. Sentencing Comm’n 2017) 277 Advisory Sentencing Guidelines 44 – 45 (Mass. Sentencing Comm’n 2017), http:// [].
Open-Ended Discretion Partially Mandatory Consecutive
MI Mich. Comp. Laws Ann. § 768.7b (West 2000) People v. Leal, 248 N.W.2d 252, 253 (Mich. Ct. App. 1976) Mandatory Concurrent Partially Mandatory Consecutive
MN Minn. Stat. §§ 609.035(b)–(c), 609.15, 609.485, 609.2232 (2017) Minn. Sentencing Guidelines and Commentary § 2F (Minn. Sentencing Guidelines Comm’n 2017) 278 Minn. Sentencing Guidelines and Commentary § 2F (Minn. Sentencing Guidelines Comm’n 2017),
Open-Ended Discretion
(Default Concurrent)
MS Miss. R. Crim. P. 26.7 Miss. Code Ann. §§ 97-3-101, 97-44-19, 99-19-21 (2014) Open-Ended Discretion
(Default Concurrent)
Partially Mandatory Consecutive
MO Mo. Sup. Ct. R. 29.09 Mo. Rev. Stat. § 558.026 (2012) Open-Ended Discretion
(Default Concurrent)
Partially Mandatory Consecutive
MT Mont. Code Ann. §§ 46-18-221, 46-18-401, 46-18-502 (2017) Open-Ended Discretion
(Default Consecutive)
Partially Mandatory Consecutive
NE Neb. Rev. Stat. §§ 28-105, 28-932, 28-1205, 29-2204 (2016) Open-Ended Discretion Partially Mandatory Consecutive
NV Nev. Rev. Stat. §§ 176.035, 193.165 (2015) Open-Ended Discretion
(Default Concurrent)
Partially Mandatory Consecutive
NH N.H. Rev. Stat. Ann. §§ 263:94, 597:14-b, 642:8, 642:9, 651:2, 651:3 (2014) Duquette v. Warden, 919 A.2d 767, 770–73 (N.H. 2007) Open-Ended Discretion 279 As noted in the cited opinion, New Hampshire had a statute in place until 1975 that required “all sentences to run concurrently except those of persons convicted of a felony either during imprisonment or an escape from imprisonment.” Duquette, 919 A.2d at 771–73 (holding that in repealing that statute, the legislature restored “the common law authority of judges to impose consecutive sentences”). Partially Mandatory Consecutive
NJ N.J. Stat. Ann. §§ 2C:11-5.2(d), 2C:12-1.1, 2C:12-13, 2C:13-1, 2C:44-5 (West 2015) State v. Miller, 13 A.3d 873, 884 (N.J. 2011) Structured Discretion Both
NM N.M. Stat. Ann. § 31-18-21(LexisNexis 2000) State v. Lopez, 661 P.2d 890, 891 (N.M. Ct. App. 1983) Open-Ended Discretion Partially Mandatory Consecutive
NY N.Y. Penal Law §§ 70.25, 70.30 (McKinney 2009) Open-Ended Discretion
(Default Concurrent)
NC N.C. Gen. Stat. §§ 14-7.12, 14-7.41(b), 15A-1340.15, 15A-1340.16E, 15A-1354(a), 15A-1354(b)(1) (2017) Open-Ended Discretion
(Default Concurrent)
ND N.D. Cent. Code § 12.1-32-11 (2012) Open-Ended Discretion
(Default Concurrent)
Partially Mandatory Concurrent
OH Ohio Rev. Code Ann. §§ 2929.14 (C)(4), 2929.41 (West Supp. 2017) State v. Comer, 793 N.E.2d 473, 477 (Ohio 2003) Felony Sentencing Quick Reference Guide 4 (Ohio Criminal Sentencing Comm’n 2015) 280 Felony Sentencing Quick Reference Guide 4 (Ohio Criminal Sentencing Comm’n 2015),
felonyQuickRef.pdf [].
Structured Discretion 281 Ohio requires that “a prison term, jail term, or sentence of imprisonment shall be served concurrently,” Ohio Rev. Code Ann. §2929.41(A), unless “the court finds that . . . consecutive service is necessary to protect the public from future crime or to punish the offender” after considering a series of enumerated factors, id. §2929.14(C)(4). Both
OK Okla. Stat. Ann. tit. 22, § 976 (1951) Open-Ended Discretion n/a
OR Or. Rev. Stat. §§ 137.121, 137.123, 137.370 (2015) Or. Admin. R. 213-012-0020 (2018) Structured Discretion 282 Oregon permits consecutive sentences to be imposed if certain statutory criteria are satisfied. See Or. Rev. Stat. § 137.123(5). If imposed, consecutive sentences are also subject to a cap that functions as a partial-concurrent-sentencing rule. See id. §137.121 (“[T]he maximum consecutive sentences which may be imposed . . . shall be as provided by rules of the Oregon Criminal Justice Commission.”); Or. Admin. R. 213-012-0020 (setting cap). Both
PA Pa. R. Crim. P. 705 18 Pa. Stat. and Cons. Stat. Ann. § 3502 (West 2013); 42 Pa. Stat. and Cons. Stat. Ann. §§ 9711.1, 9721, 9765 (West 2013); 75 Pa. Stat. and Cons. Stat. Ann. § 3735 (West 2013) Open-Ended Discretion Both
RI 11 R.I. Gen. Laws § 11-47-3.2 (Supp. 2016); 11 R.I. Gen. Laws § 11-25-2 (2002); 12 id. § 12-19-5 1 R.I. Ct. R. Ann., Super. Ct. Sentencing Benchmarks: Using the Benchmarks § 7, at 684 (Lexis 2017) Open-Ended Discretion Partially Mandatory Consecutive
SC S.C. Code Ann. § 16-15-395 (2015) State v. Barton, 481 S.E.2d 439, 444 (S.C. Ct. App. 1997) Open-Ended Discretion Partially Mandatory Consecutive
SD S.D. Codified Laws § 22-6-6.1 (2006) Open-Ended Discretion 283 Sentences are by default consecutive if the defendant commits a crime while incarcerated. S.D. Codified Laws § 23A-27-36.1. n/a
TN Tenn. R. Crim. P. 32 Tenn. Code Ann. §§ 40-20-111, 40-35-115 (2014) Structured Discretion Partially Mandatory Consecutive
TX Tex. Penal Code § 3.03 (2017); Tex. Code Crim. Proc. art. 42.08 (2017) Mandatory Concurrent Both
UT Utah Code Ann. §§ 58-37-8, 76-3-401(LexisNexis 2016) Adult Sentencing & Release Guidelines 20 (Utah Sentencing Comm’n 2017) 284 Adult Sentencing & Release Guidelines 20 (Utah Sentencing Comm’n 2017), []. The Utah Sentencing Guidelines impose a unique hybrid regime that is simultaneously mandatory consecutive and mandatory concurrent. Specifically, “[i]f multiple convictions are ordered to run concurrently, the guidelines add 10% of the recommended length of stay of the shorter sentence to the full recommended length of the longer sentence.” Id. At the same time, “[i]f multiple convictions are ordered to run consecutively, the guidelines add 40% of the recommended length of stay of the shorter sentence to the full recommended length of the longer sentence.” Id. As the guidelines further explain, this means that a judge considering imposing two sentences at once, one for 84 months the other for 20, can either sentence the defendant to a “concurrent” term of 86 months or a “consecutive” term of 92 months. Id. Judges in Utah “are encouraged to sentence within the guidelines.” Id. at 12. Structured Discretion Both
VT Vt. Stat. Ann. tit. 13, §§ 1028, 1501, 5409, 7032 (2009) Open-Ended Discretion Partially Mandatory Consecutive
VA Va. Code Ann. § 19.2-308 (2015); Va. Code Ann. §§ 18.2-53.1, 18.2-255.2, 18.2-308.1 (2014) Open-Ended Discretion
(Default Consecutive)
Partially Mandatory Consecutive
WA Wash. Rev. Code §§ 9.92.080(2), 9.94A.535, 9.94A.589 (2017)


Washington State Adult Sentencing Guidelines Manual 30–32 (Wash. State Caseload Forecast Council 2016) 285 Washington State Adult Sentencing Guidelines Manual 30–32 (Wash. State Caseload Forecast Council 2016),
Manual_2016.pdf [].
Structured Discretion Partially Mandatory Consecutive
WV W. Va. Code Ann. §§ 61-2-10b(f), 61-11-21, 62-8-2(c) (LexisNexis 2014) Open-Ended Discretion
(Default Consecutive)
Partially Mandatory Consecutive
WI Wis. Stat. §§ 973.10, 973.15 (2018) Open-Ended Discretion n/a
WY Wyo. R. Crim. P. 32 Open-Ended Discretion n/a
FED 18 U.S.C. §§ 924(c), 1028A, 2250, 3553(a), 3584 (2012) U.S. Sentencing Guidelines Manual § 3D1.1–.5 (U.S. Sentencing Comm’n 2016) 286 U.S. Sentencing Guidelines Manual § 3D1.1–.5 (U.S. Sentencing Comm’n 2016), [http://].
Structured Discretion Both

Table C: The Law of Pretrial Evidentiary Review§

§ tabThis table supports Figure 14, supra (p. 1346). The law of pretrial evidentiary review is the most complicated body of procedural law addressed in this Article. In a number of states, a conglomeration of rules, statutes, constitutional provisions, and judicial opinions all come together to construct the governing regime. Accordingly, the table reports in its rightmost column the substantive code for each state’s procedural framework, as those codes are described in section II.A. See supra text accompanying notes 109–124. The supporting authorities are then organized by lawmaking modality, with parenthetical codes describing which component of the regime the corresponding authority establishes. Provisions that pertain to the existence or nonexistence of a preliminary hearing are coded “hearing”; provisions that pertain to the existence or nonexistence of papers review are coded “papers review”; provisions that pertain to the existence or nonexistence of hearing bypass (that is, indictment bypass or information bypass) are coded “bypass.” Where a state authorizes both indictment bypass and information bypass, it is coded as the latter. See supra text accompanying notes 114–117 (explaining why prosecutors will generally elect to use information-bypass authority if it is available to them).

State Source of Authority Procedural Regime
Court Rule Statute Constitution Case Law
AL Ala. R. Crim. P. 5.1(a) (hearing, bypass); id. 12.8 (papers review) Ala. Code § 15-11-1 (LexisNexis 2015) (hearing, bypass) Indictment Bypass with Papers Review
AK Alaska R. Crim. P. 5(e)(bypass) Nelson v. State, 628 P.2d 884, 890 (Alaska 1981) (papers review) Indictment Bypass with Papers Review
AZ Ariz. R. Crim. P. 5.1 & cmt. (bypass); id. 12.9(a) (papers review) 287 Rule 12.9(a) permits a defendant to “fil[e] a motion for a new finding of probable cause . . . [if] the defendant was denied a substantial procedural right.” Ariz. R. Crim. P. 12.9(a). However, if “a duly constituted grand jury returns an indictment valid on its face and a challenge is made to the kind of evidence considered by the grand jury in making its decision, [then] Rule 12.9(a) does not apply.” State v. Superior Court, 577 P.2d 743, 745 (Ariz. Ct. App. 1978) (citation omitted). Ariz. Rev. Stat. Ann. § 13-3951 (2010) (hearing) Ariz. Const. art. 2, § 30 (hearing) Indictment Bypass
AR Ark. R. Crim. P. 8.3(c) (hearing)


Ark. Code Ann. § 16-85-302 (2005) (hearing); id. §16-85-302 (bypass); id. § 16-85-706 (papers review) 288 See also Ware v. State, 75 S.W.3d 165, 170 (Ark. 2002) (“Lack of probable cause is not a statutory ground for a motion to set aside an indictment or, by implication, to quash an information.”). No Review
CA Cal. Penal Code § 859b (2018) (hearing); id. § 738 (bypass, papers review); id. §§ 939.71, 939.6(b), 995 (papers review) Cal. Const. art. 1, §§ 14–14.1 (hearing, bypass) Indictment Bypass with Papers Review
CO Colo. R. Crim. P. 5(a)(4) (hearing); id. 7(c) (bypass)


Colo. Rev. Stat. § 16-5-301 (2017) (hearing, bypass); id. §§ 16-5-204(k), 16-5-205 (papers review) Indictment Bypass with Papers Review
CT Conn. Super. Ct. R. § 41-8(5) (papers review) Conn. Gen. Stat. Ann. § 54-46a(a) (West 2017) (hearing); id. § 54-56 (papers review) Conn. Const. art. 1, § 8 (hearing)


Papers Review 289 Preliminary hearings are authorized only in cases in which the defendant is accused of a “crime punishable by death, life imprisonment without the possibility of release or life imprisonment.” Conn. Gen. Stat. Ann. § 54-46a; cf. State v. Rollinson, 526 A.2d 1283, 1286 (Conn. 1987) (explaining the “statute was enacted to implement the provisions of article XVII of the amendments to the Connecticut constitution, which substituted a probable cause hearing for the grand jury indictment formerly required before a person could be held to answer for a crime punishable by death or life imprisonment”). Papers review is available outside of the relatively narrow class of cases in which a preliminary hearing exists.
DE Del. Super. Ct. R. Crim. P. 5 (hearing, bypass)


Del. Code Ann. tit. 11, § 5308 (2018) (hearing)


Ellegood v. State, 782 A.2d 263, 2001 WL 770264, at *2 (Del. June 28, 2001) (unpublished table decision) (papers review) 290 The cited case establishes that papers review is not available for an indictment. Delaware similarly does not appear to have any provision affording papers review for an information. Both an indictment and an information can bypass a preliminary hearing. See Del. Super. Ct. R. Crim. P. 5(d). Information Bypass
FL Fla. R. Crim. P. 3.133(b) (hearing, bypass); id. 3.190(c)(4) (papers review) 291 While papers review is available as a means to challenge an information, Florida courts may “never inquire into the character of the evidence that influenced a grand jury in finding [an] indictment.” State v. Schroeder, 112 So. 2d 257, 260 (Fla. 1959) (emphasis omitted) (quoting Richardson v. State, 130 So. 718, 720 (Fla. 1930)). Information Bypass with Papers Review
GA Ga. Super. Ct. Unif. R. 26.1–.2; Ga. Magis. Ct. Unif. R. 25 (hearing, bypass) Ga. Code Ann. § 17-7-23 (2013) (hearing) State v. Middlebrooks, 222 S.E.2d 343, 345–46 (Ga. 1976) (bypass); Fitzgerald v. State, 304 S.E.2d 114, 116 (Ga. Ct. App. 1983) (papers review) Indictment Bypass
HI Haw. R. Penal P. 5(c) (hearing, bypass) Haw. Rev. Stat. Ann. §§ 806-85 to -87 (LexisNexis 2016) (papers review) Haw. Const. art. I, § 10 (hearing, bypass) State v. Ontai, 929 P.2d 69, 76 (Haw. 1996) (papers review) Indictment Bypass with Papers Review
ID Idaho Crim. R. 5.1(a) (hearing, bypass) Idaho Code §§ 19-804, 19-1308 (2017) (hearing) 292 Idaho affords defendants a form of papers review of the preliminary hearing itself, insofar as it permits them to “challenge the sufficiency of evidence educed at the preliminary examination” by seeking review before a district court judge. Idaho Code § 19-815A. State v. Brandstetter, 908 P.2d 578, 580 (Idaho Ct. App. 1995) (papers review) Indictment Bypass with Papers Review
IL 725 Ill. Comp. Stat. Ann. 5/109-3.1(b), 5/111-2 (West 2006) (hearing, bypass) Ill. Const. art. I, § 7 (hearing, bypass) People v. Bragg, 467 N.E.2d 1004, 1008 (Ill. App. Ct. 1984) (papers review) Indictment Bypass 293 A statutory provision permits an indictment to be challenged on the narrow ground that it “is based solely upon the testimony of an incompetent witness.” 725 Ill. Comp. Stat. Ann. 5/114-1(a)(9). However, case law makes clear that, notwithstanding that narrow form of papers review, “[a] defendant may not challenge an indictment on the grounds that it is not supported by adequate evidence.” Bragg, 467 N.E.2d at 1008.
IN Ind. Code §§ 35-33-7-1, 35-33-7-2 (2017) (hearing); id. §35-34-1-4 (papers review) No Review 294 See also Schutz v. State, 413 N.E.2d 913, 916 (Ind. 1981) (“We have long held that an indictment or information may not be questioned on the ground of insufficient evidence. The sufficiency of the evidence is decided at trial.”).
IA Iowa Ct. R. 2.2(4) (hearing, bypass); id. 2.5(4) (papers review)  


Information Bypass with Papers Review
KS Kan. Stat. Ann. § 22-2902(1) (2007) (hearing, bypass) Indictment Bypass 295 Note that in misdemeanor cases, “a prosecution may be begun by filing an information . . . only after the judge has determined from the information . . . or from other evidence that there is probable cause to believe both that a crime has been committed and that the defendant has committed it.” Kan. Stat. Ann. § 22-2303(1).
KY Ky. R. Crim. P. 3.07 (hearing, bypass); id. 5.10 (papers review) Indictment Bypass
LA La. Code Crim. Proc. Ann. art. 292, 296 (2003) (hearing, bypass); id. art. 532 (papers review) La. Const. art. I, § 14 (hearing, bypass) Indictment Bypass
ME State v. Rizzo, 704 A.2d 339, 342 (Me. 1997) (papers review) No Review 296 Maine repealed its statutory provisions providing for preliminary hearings in 1965. See Act Effective Dec. 1, 1965, 1965 Me. Laws 455.
MD Md. R. 4-221(c)(1) (hearing, bypass) Md. Code Ann., Crim. Proc. §§ 4-102, 4-103(c)(1) (LexisNexis 2008) (hearing, bypass) State v. Taylor, 810 A.2d 964, 981 (Md. 2002) (papers review) Indictment Bypass 297 A Maryland statute purports to allow a preliminary hearing to be held at the court’s discretion, see Md. Code Ann., Crim. Proc. § 4-103(c)(2), but a countervailing rule does not allow such discretion and takes priority. See Marshall v. State, 420 A.2d 1266, 1269 (Md. Ct. Spec. App. 1980), rev’d on other grounds, 434 A.2d 555 (Md. 1981).
MA Mass. R. Crim. P. 3(e)–(f) (hearing, bypass); id. 4(c) & cmt. (papers review) Mass. Ann. Laws ch. 276, § 38 (LexisNexis 2010) (hearing); id. ch. 263, § 4A (bypass) Commonwealth v. Capone, 659 N.E.2d 1196, 1198–99 (Mass. App. Ct. 1996) (papers review) Indictment Bypass with Papers Review
MI Mich. R. Crim. P. 6.110, 6.112(B) (hearing, bypass) Mich. Comp. Laws Ann. § 766.1 (West 2000) (hearing) People v. Auto. Serv. Councils of Mich., Inc., No. 492, 1980 WL 4690, at *20 (Mich. Dist. Sept. 5, 1980), aff’d, 333 N.W.2d 352 (Mich. Ct. App. 1983) (papers review) Indictment Bypass with Papers Review 298 See also Mich. Comp. Laws Ann. § 767.42 (“An information shall not be filed against any person for a felony until such person has had a preliminary examination therefor . . . .”).
MN Minn. R. Crim. P. 11.04 & cmt.; id. 18.06 (papers review) Papers Review 299 See State v. Florence, 239 N.W.2d 892, 902 (Minn. 1976) (discussing papers review); State v. Eibensteiner, 690 N.W.2d 140, 150–52 (Minn. Ct. App. 2004) (same).
MS Miss. R. Crim. P. 6.1(a)(1) (hearing, bypass) State v. Matthews, 218 So. 2d 743, 744 (Miss. 1969) (papers review) Indictment Bypass
MO Mo. Sup. Ct. R. 22.09 (hearing) Mo. Rev. Stat. § 544.250 (2017) (hearing) Mo. Const. art. I, § 17 (hearing) State v. Green, 305 S.W.2d 863, 868 (Mo. 1957) (bypass); State v. Burkhart, 615 S.W.2d 565, 568 (Mo. Ct. App. 1981) (papers review) Indictment Bypass
MT Mont. Code. Ann. § 46-10-105 (2017) (hearing, bypass); id. §§ 46-11-201, 46-11-203 (papers review) Mont. Const. art. 2, § 20 (hearing) Information Bypass with Papers Review 300 Montana also permits bypass by indictment. See Mont. Code. Ann. 46-11-101.
NE Neb. Rev. Stat. §§ 29-504, 29-1607 (2016) (hearing, bypass); id. § 29-1418 (grand jury papers review) Hearing as of Right with Papers Review 301 See supra note 132 (discussing Nebraska provisions in detail); see also Neb. Rev. Stat § 29-1607 (“No information shall be filed against any person for any offense until such person shall have had a preliminary examination therefor . . . .”).
NV Nev. Rev. Stat. §§ 171.186, 171.202 (2015) (hearing) Cairns v. Sheriff, 508 P.2d 1015, 1017 (Nev. 1973) (bypass); Sheriff v. Miley, 663 P.2d 343, 344 (Nev. 1983) (papers review) Indictment Bypass with Papers Review 302 See also Nev. Rev. Stat. § 173.035(1) (providing that “[a]n information may be filed against any person for any offense when the person . . . [h]as had a preliminary examination”).
NH N.H. R. Crim. P. 6(a) (hearing, bypass) N.H. Rev. Stat. Ann. § 592-B:7 (2015) (hearing, bypass) State v. Williams, 708 A.2d 55, 57 (N.H. 1998) (papers review) Indictment Bypass
NJ N.J. Ct. R. 3:4-3(a) (hearing, bypass) N.J. Const. art. 1, ¶ 8 (hearing, bypass) State v. Hogan, 676 A.2d 533, 543 (N.J. 1996) (papers review) Indictment Bypass
NM N.M. R. 6-202(a), (f) (hearing, bypass) N.M. Stat. Ann. § 31-6-11(A) (LexisNexis 2000) (papers review) N.M. Const. art. II, § 14 (hearing, bypass) Indictment Bypass
NY N.Y. Crim. Proc. Law § 180.10(2) (McKinney 2007) (hearing, bypass); id. §§ 210.20(b), 210.30 (papers review) N.Y. Const. art. 1, § 6 (hearing) Indictment Bypass with Papers Review
NC N.C. Gen. Stat. § 15A-606 (2017) (hearing); id. § 15A-955 (papers review) State v. Lester, 240 S.E.2d 391, 395 (1978) (bypass) Indictment Bypass
ND N.D. R. Crim. P. 5(c) (hearing) Hearing as of Right 303 The North Dakota Supreme Court has suggested that, “[i]n effect, a preliminary examination is in lieu of the grand jury proceedings and indictment,” language that suggests an indictment-bypass regime. Dickinson Newspapers, Inc. v. Jorgensen, 338 N.W.2d 72, 75 (N.D. 1983). The same opinion, however, states that “[t]he provisions for preliminary examination (hearing) . . . contained in Rules 5(c) and 5.1 of the North Dakota Rules of Criminal Procedure . . . basically provide that a person charged with a felony has a right to a preliminary examination . . . .” Id. Additional authority supports the hearing-as-of-right classification. See N.D. R. Crim. P. 5(c) (“If the offense charged is a felony, the defendant has the right to a preliminary hearing.”); Nick Thornton, What Happens in a Criminal Case?, Fremstad Law (Nov. 26, 2014), [] (“In a felony case, the defendant has the right to a preliminary hearing.”). But cf. State v. Nordquist, 309 N.W.2d 109, 116 (N.D. 1981) (conducting papers review of an indictment without mention of a preliminary hearing).
OH Ohio R. Crim. P. 5(A)(4), (B)(1) (hearing, bypass); id. 12(c)(2), 48 cmt. (papers review) Ohio Rev. Code Ann. § 2937.02 (West 2006) (hearing); id. § 2937.04 (papers review) Indictment Bypass
OK Okla. Stat. Ann. tit. 22, §§ 253, 258, 524 (West 2003) (hearing, bypass); id. § 504.1 (papers review) Okla. Const. art. II, § 17 (hearing) Hearing as of Right with Papers Review 304 See Stone v. Hope, 488 P.2d 616, 618–19 (Okla. Crim. App. 1971) (“Under our criminal procedure after the filing of an Information . . . the accused is entitled to a preliminary examination . . . . Prosecution by Indictment likewise entitles [the] accused to a preliminary hearing.”); see also State v. Delso, 298 P.3d 1192, 1193–94 (Okla. Crim. App. 2013) (“[A] defendant [can] file a motion to quash for insufficient evidence in felony cases after preliminary hearing [if he can] establish . . . that there is insufficient evidence to prove any one of the necessary elements of the offense for which the defendant is charged.” (citing Okla. Stat. tit. 22, § 504.1(A))).
OR Or. Rev. Stat. § 135.070(2) (2015) (hearing); id. § 135.510 (papers review) Or. Const. art. VII (amended), § 5 (hearing) State v. Marsh, 888 P.2d 580, 582 (Or. Ct. App. 1995) (bypass) Indictment Bypass
PA Pa. R. Crim. P. 540(F)(2), 541 (hearing); id. 556, 556.2 & cmt. (bypass); id. 556.4(b)(1)(B) (papers review) Hearing as of Right with Papers Review 305 See supra note 132 (discussing narrow exceptions to hearing as of right and supplemental papers review). In addition to the exceptions noted earlier, “the court may grant leave to the Commonwealth to file an information with the court without a preliminary hearing when the district attorney certifies that a preliminary hearing cannot be held for good cause shown.” Commonwealth v. Cassidy, 620 A.2d 9, 11 (Pa. Super. Ct. 1993); see also Pa. R. Crim. P. 565.
RI R.I. Dist. Ct. R. Crim. P. 5(c) (hearing); R.I. Super. Ct. R. Crim. P. 9.1 (papers review) 12 R.I. Gen. Laws § 12-10-5 (2002) (hearing); id. § 12-12-1.7 (papers review) Information Bypass with Papers Review 306 The rule and statutory provisions cited indicate that a preliminary hearing is available only when a defendant is facing a criminal complaint in the district court, not when the defendant is facing an information, which prosecutors are free to file as the initial charging document. It would appear from this framework that preliminary hearings exist at the prosecutor’s option, as in a typical information-bypass regime, and thus rarely occur. And indeed, an overview of the Rhode Island criminal process published by the state’s judiciary makes no reference to preliminary hearings (though it does reference bail hearings for defendants held without bail). See Criminal Case Process, R.I. Judiciary, [
ZZ69-J666] (last visited April 4, 2018). A defendant may, however, move to dismiss an information, in which case the court “is required to examine the information and any attached exhibits to determine whether the state has satisfied its burden to establish probable cause.” State v. Fritz, 801 A.2d 679, 682 (R.I. 2002) (citing State v. Aponte, 649 A.2d 219, 222 (R.I. 1994) (per curiam)).
SC S.C. R. Crim. P. 2(a)–(b) (hearing, bypass) S.C. Code Ann. § 17-23-160 (2014) (hearing); id. § 22-5-320 (2007) (hearing) State v. Williams, 210 S.E.2d 298, 301 (S.C. 1974) (papers review) Indictment Bypass
SD S.D. R. Crim. P. 5(c) (hearing) S.D. Codified Laws § 23A-4-3 (2016) (hearing, bypass) State v. Carothers, 724 N.W.2d 610, 616 (S.D. 2006) (papers review) Indictment Bypass
TN Tenn. R. Crim. P. 5(e) (hearing, bypass) Hearing as of Right
TX Tex. Code Crim. Proc. art. 16.01 (2017) (hearing, bypass) DeLeon v. State, 758 S.W.2d 621, 623 (Tex. Ct. App. 1988) (bypass); Dean v. State, 749 S.W.2d 80, 81 (Tex. Crim. App. 1988) (en banc) (papers review) Indictment Bypass 307 While the statutory text suggests that the defendant “shall have the right to an examining trial before indictment,” Tex. Code Crim. Proc. art. 16.01, Texas courts have consistently held that “[t]he return of an indictment terminates the right to an examining trial.” DeLeon, 758 S.W.2d at 623.
UT Utah R. Crim. P. 7(e) (hearing, bypass) Utah Const. art. 1, § 13 (hearing, bypass) Indictment Bypass with Papers Review 308 Utah’s indictment-bypass system is unusual: Before a case can be presented to a grand jury, a panel of five judges must first review the evidence; it is those judges who then authorize the case to proceed to a grand jury. See Utah Code Ann. § 77-10a-2(2)(a) (LexisNexis 2012) (“If the panel finds good cause to believe a grand jury is necessary, the panel shall make its findings in writing and may order a grand jury to be summoned.”). If a grand jury is then empanelled and returns an indictment, the indictment bypasses a preliminary hearing. Thus, Utah allows what is essentially an ex parte preliminary hearing (before the panel of five judges) to bypass a more traditional preliminary hearing—and in so doing affords a more robust check on the indictment process than the somewhat imprecise “papers review” code in the table reflects. Cf. id. § 77-10a-2(1)(c) (noting that “[h]earsay evidence may be presented at the [five-judge] hearings only under the same provisions and limitations that apply to preliminary hearings”). Alternatively, if the prosecutor does not want to go before the five-judge panel, she may instead file charges via information, in which case the defendant will have a right to a preliminary hearing. See Utah Const. art. 1, §13.
VT Vt. R. Crim. P. 5(c) (hearing, bypass); id. 12(d) & reporter’s note to 2008 amendment (papers review) Indictment Bypass with Papers Review
VA Va. Code Ann. §§ 19.2-183, 19.2-218 (2015) (hearing, bypass) Indictment Bypass 309 See Seibert v. Commonwealth, 467 S.E.2d 838, 840 (Va. Ct. App. 1996) (carving out a significant exception to statutory text that would otherwise seem to grant preliminary hearings as of right); see also supra note 132.
WA Wash. Crim. R. Cts. Limited Jurisdiction 3.2.1(g) (hearing, bypass) State v. Knapstad, 729 P.2d 48, 51 (Wash. 1986) (papers review) Information Bypass with Papers Review
WV W. Va. R. Crim. P. 5(c), 7 (hearing, bypass) W. Va. Code Ann. § 62-1-8 (LexisNexis 2014) (hearing) State v. Davis, 782 S.E.2d 423, 428 (W. Va. 2015) (bypass); Barker v. Fox, 238 S.E.2d 235, 236 (W. Va. 1977) (papers review) Indictment Bypass 310 While the governing rule suggests that the defendant has “the right to a preliminary examination,” W. Va. R. Crim. P. 5(c), the West Virginia Supreme Court has made clear that “a preliminary hearing is not required if the state elects to proceed directly to a grand jury without arresting the accused, or if the accused is arrested but indicted before the date set for the preliminary examination.” Davis, 782 S.E.2d at 428. The Rule also states “that the preliminary examination shall not be held . . . if an information against the defendant is filed in circuit court before the date set for the preliminary examination.” W. Va. R. Crim. P. 5(c). However, a “felony offense may be prosecuted by information [only] if the indictment is waived” by the defendant, W. Va. R. Crim. P. 7(a), which renders West Virginia an indictment-bypass regime.
WI Wis. Stat. §§ 968.06, 970.02(1)(c), 971.02(1) (2018) (hearing, bypass) Hearing as of Right
WY Wyo. R. Crim. P. 5.1(a) (hearing, bypass) Wyo. Stat. Ann. § 7-8-105 (2017) (hearing, bypass) Hennigan v. State, 746 P.2d 360, 371 (Wyo. 1987) (papers review) Indictment Bypass
FED Fed. R. Crim. P. 5.1 (hearing, bypass) Costello v. United States, 350 U.S. 359, 363 (1956) (papers review) Indictment Bypass


Table D: The Law of Bills of Particulars||

|| This table supports Figure 15, supra (p. 1356), and collects authorities governing bills of particulars across the states. The focus here is on whether the defendant has either the opportunity or the right to obtain a judicial order requiring the prosecutor to provide a specific statement of the factual basis underlying the charges. “Specific” here is intended to refer to information that goes beyond the time or location of the offense and to capture instead factual detail sufficient to permit the defendant to mount a legal challenge to a concrete factual theory of the case. See supra text accompanying notes 156–164 (discussing the connection between factual specificity, legal overreach, and motions for summary dismissal). Where the defendant is entitled to such a statement, the state is coded as “mandatory.” Alternatively, the “discretionary” code is applied to those states in which the trial court has discretion to order the prosecutor to produce such a statement—either because the governing law is stated in expressly discretionary terms (for example, “the court may order”) or because it employs an open-ended standard that is essentially discretionary in application. Cf. 5 LaFave et al., supra note 73, § 19.4(b), at 361 (“In several states, statutory provisions make issuance of particulars mandatory, but that obligation tends to be tied to broadly stated standards that give the trial judge considerable leeway.” (footnote omitted)). Finally, a number of states have abolished bills of particulars altogether. Note, however, that the factual specificity potentially afforded by a bill of particulars can also be guaranteed in other ways, including by a requirement that the charging document itself state the allegations with factual specificity. Cf. Nance v. State, 918 S.W.2d 114, 124 (Ark. 1996) (“[W]here the information is definite in specifying the offense being charged, as in this case, the charge itself constitutes a bill of particulars.”). Accordingly, the “not available” code is applied here only to those states in which the bill of particulars does not exist and in which the corresponding charging rules also do not require factual specificity (as defined above).

State Source of Authority Procedural Regime
Court Rule Statute Case Law
AL Ala. R. Crim. P. 13.2(e) Discretionary 311 “A defendant is not entitled to a bill of particulars under Alabama law.” McCrary v. State, 398 So. 2d 752, 756 (Ala. Crim. App. 1981); see also Jones v. State, 34 So. 236, 238 (Ala. 1903). However, commentary to the rule cited above provides “that for good cause shown the defendant can compel the state to submit additional details of the offense not required to be set out in the body of the indictment.” Ala. R. Crim. P. 13.2 cmt. As the Rule Committee further explains, “it is contemplated that motions for more definite statement will not be routinely made or granted” given “the ‘good cause’ requirement.” Id.
AK Alaska R. Crim. P. 7(f) Discretionary
AZ State v. Schwartz, 484 P.2d 1060, 1062 (Ariz. Ct. App. 1971) Discretionary
AR Ark. Code Ann. §§ 16-85-301, 16-85-403(a)(2) (2005) Discretionary 312 The governing statute provides that, “Upon request of the defendant, the state shall file a bill of particulars setting out the act or acts upon which it relies for conviction.” Ark. Code Ann. § 16-85-403(a)(2). However, case law indicates that “[t]he trial court, using discretion, can grant or deny the request.” Burnett v. State, 697 S.W.2d 95, 99 (Ark. 1985), overruled on other grounds by Midgett v. State, 729 S.W.2d 410 (Ark. 1987).
CA People v. Thorn, 33 P.2d 5, 15 (Cal. Dist. Ct. App. 1934) Not Available 313 See also Cal. Penal Code § 952 (2018) (permitting charges to be filed without factual specificity).
CO Colo. R. Crim. P. 7(g) Discretionary 314 While the governing rule uses expressly discretionary language, case law indicates that a bill of particulars is mandatory in certain limited cases. See People v. Dist. Court for the Second Judicial Dist., 603 P.2d 127, 128 (Colo. 1979) (en banc) (“An order for a bill of particulars normally rests within the sound discretion of the trial judge. Where the crime of theft is charged in the words of the statute, however, such an order is mandatory upon the defendant’s request.” (citations omitted)).
CT Conn. R. Super. Ct. §§ 41-20 to -22 Discretionary
DE Del. Super. Ct. R. Crim. P. 7(f) Discretionary
FL Fla. R. Crim. P. 3.140(n) Discretionary 315 Florida is a good example of a state that combines a putatively mandatory issuance requirement with an open-ended standard, thereby producing a de facto discretionary regime. Compare Fla. R. Crim. P. 3.140(n) (“The court, on motion, shall order the prosecuting attorney to furnish a statement of particulars when the indictment or information on which the defendant is to be tried fails to inform the defendant of the particulars of the offense sufficiently to enable the defendant to prepare a defense.” (emphasis added)), with Saldana v. State, 980 So. 2d 1220, 1222 (Fla. Dist. Ct. App. 2008) (“We . . . conclude that the trial court did not abuse its discretion in denying [the defendant’s] motion for a statement of particulars . . . .”), and Peel v. State, 154 So. 2d 910, 912 (Fla. Dist. Ct. App. 1963) (“[A] bill of particulars is never required in a criminal case in Florida except in exceptional cases where the denial of same constitutes an abuse of judicial discretion.”).
GA Freeman v. State, 392 S.E.2d 330, 332 (Ga. Ct. App. 1990) Not Available 316 See also Ga. Code Ann. § 17-7-54 (2013) (permitting charges to be filed without factual specificity).
HI Haw. R. Penal P. 7(g) Haw. Rev. Stat. Ann. § 806-47 (LexisNexis 2016) Discretionary
ID State v. Holcomb, 912 P.2d 664, 669 (Idaho Ct. App. 1995) Discretionary 317 In the cited case, the court “[a]ssum[ed], without deciding,” that “orders for bills of particulars are still permissible” under the common law even though “there is no provision for bills of particulars in the Idaho Criminal Rules.” Holcomb, 912 P.2d at 669.
IL 725 Ill. Comp. Stat. Ann. 5/111-6 (West 2006) Discretionary
IN Sherrick v. State, 79 N.E. 193, 194 (Ind. 1906) Not Available 318 In the cited case, the Indiana Supreme Court made clear that “there is no recognition of a motion for a bill of particulars” under the state’s law, but went on to suggest “that, under the certainty required in criminal pleading in this state, whenever a trial judge finds it necessary to the administration of justice to grant a bill of particulars, he has found an ample reason for quashing the indictment for uncertainty.” Sherrick, 79 N.E. at 194. This language suggests that a robust factual-specificity requirement at the charging stage may at one point in time have offset the state’s nonrecognition of bills of particulars. The current charging statute, however, does not require any factual specificity. See Ind. Code. § 35-34-1-2(a), (e) (2017).
IA Iowa Ct. R. 2.11(5) Discretionary
KS Kan. Stat. Ann. § 22-3201(f) (2007) Discretionary
KY Ky. R. Crim. P. 6.22 Discretionary
LA La. Code Crim. Proc. Ann. art. 484 (2017) Discretionary
ME Me. R. Unified Crim. P. 16(d)(1) Discretionary
MD Md. R. 4-241 Discretionary
MA Mass. R. Crim. P. 13(b) Discretionary
MI Mich. R. Crim. P. 6.112(E) Discretionary
MN Minn. R. Crim. P. 17.02(4) Not Available 319 The Minnesota Supreme Court amended Rule 17.02 in 2008 so as to eliminate bills of particulars. See Order, Promulgation of Amendments to the Minnesota Rules of Criminal Procedure, No. C1-84-2137, at 7 (Minn. 2008),
mncourtsgov/media/AdministrativeFileArchive/Criminal%20Procedure%20Rules%20ADM10-8049%20formerly%20C1-84-2137/2008-05-19-Order-Crim-Proc-Amendments.pdf [http://] (“The bill of particulars is abolished.”).
MS Tapper v. State, 47 So. 3d 95, 108 (Miss. 2010) Not Available 320 See also Smallwood v. State, 584 So. 2d 733, 738 (Miss. 1991) (permitting charges to be filed without factual specificity).
MO Mo. R. Crim. P. 23.04 Discretionary
MT Fitzpatrick v. State, 638 P.2d 1002, 1025 (Mont. 1981) Not Available 321 See also Mont. Code Ann. § 46-11-401(1) (2017) (permitting charges to be filed without factual specificity).
NE Neb. Rev. Stat. § 29-1808 (2016) State v. Case, 553 N.W.2d 173, 179–80 (Neb. Ct. App. 1996) Discretionary 322 While Nebraska does not recognize a motion for a bill of particulars, it permits a motion to quash a charging instrument that lacks factual specificity. See Neb. Rev. Stat. § 29-1808; Case, 553 N.W.2d at 180 (“To the extent that [the defendant’s] motion for bill of particulars was an effort to . . . challenge the certainty and particularity of the information for the preparation of his defense, a motion to quash is the proper method of attack.”).
NV Not Available 323 Nevada does not appear to have any provision or case law providing for bills of particulars and permits charges to be filed without factual specificity. See Nev. Rev. Stat. § 173.075 (2015); Volosin v. State, No. 72184, 2017 WL 6547445, at *1 (Nev. Ct. App. Dec. 15, 2017).
NH State v. Voorhees, 632 A.2d 825, 826 (N.H. 1993) Discretionary
NJ N.J. R. Crim. P. 3:7-5 State v. Davis, 70 A.2d 761, 762 (N.J. Super. Ct. App. Div. 1950) Discretionary
NM N.M. R. 5-205(C) Discretionary
NY N.Y. Crim. Proc. Law § 200.95(1) (McKinney 2007) Mandatory
NC N.C. Gen. Stat. § 15A-925 (2018) Discretionary
ND N.D. R. Crim. P. 7(f)


OH Ohio R. Crim. P. 7(E) Ohio Rev. Code Ann. § 2941.07 (West 2006) Mandatory
OK Okla. Stat. Ann. tit. 22, §§ 409, 504, 512 (West 2003) Not Available 324 Oklahoma permits charges to be filed without factual specificity and does not appear to provide any vehicle by which a defendant might compel such specificity from the prosecution after charges are filed.
OR State v. House, 489 P.2d 381, 382–83 (Or. 1971) Not Available
PA Pa. R. Crim. P. 572 Discretionary
RI R.I. Super. Ct. R. Crim. P. 7(f) State v. Gregson, 113 A.3d 393, 397 (R.I. 2015) Discretionary 325 The governing rule states that “the court shall direct the filing of a bill of particulars” upon the defendant’s request. R.I. Super. Ct. R. Crim. P. 7(f). But the state supreme court has held that “the granting of a bill of particulars in any civil or criminal proceeding is within the discretion of the justice who hears the motion” and has expressly upheld a trial court’s denial of a motion for such a bill when a defendant attempted to force the prosecutor to choose between two competing theories of liability. Gregson, 113 A.3d at 397 (internal quotation marks omitted) (quoting Union Mortg. Co. v. Rocheleau, 154 A. 658, 660 (R.I. 1931)).
SC State v. Wells, 161 S.E. 177, 181 (S.C. 1931) Not Available 326 See also S.C. Code Ann. § 17-19-20 (2014) (permitting charges to be filed without factual specificity).
SD State v. Anderson, 546 N.W.2d 395, 400 n.8 (S.D. 1996) Discretionary
TN Tenn. R. Crim. P. 7(c) Discretionary
TX Noel v. State, 769 S.W.2d 366, 368 (Tex. Ct. App. 1989) Discretionary 327 Texas does not use the term “bill of particulars” but does recognize a common law right to request additional factual specificity. See Noel, 769 S.W.2d at 368 (“[W]here the language concerning the defendant’s conduct is so vague or indefinite as to deny the defendant effective notice of the act he allegedly committed, the motion seeking more information should be granted.”).
UT Utah R. Crim. P. 4(e), 4A(d) Utah Code Ann. § 77-14-1 (LexisNexis 2017) State v. Allen, 839 P.2d 291, 298 (Utah 1992) Discretionary
VT State v. Bridgman, 49 Vt. 202, 206 (1876) Discretionary 328 See also State v. Sears, 296 A.2d 218, 219 (Vt. 1972) (“A bill of particulars was sought by the respondent when charged, and duly furnished.”).
VA Va. Code Ann. § 19.2-230 (2015) Discretionary 329 A separate statutory provision states in seemingly mandatory language that the “court shall, upon motion of the defendant, direct the Commonwealth to file a bill of particulars pursuant to § 19.2-230.” Va. Code Ann. § 19.2-266.2. But the state supreme court has held that so long as “an indictment . . . give[s] the accused ‘notice of the nature and character of the offense charged so he can make his defense’ . . . a bill of particulars is not required.” Strickler v. Commonwealth, 404 S.E.2d 227, 233 (Va. 1991) (quoting Wilder v. Commonwealth, 225 S.E.2d 411, 413 (Va. 1976)).
WA Wash. Super. Ct. Crim. R. 2.1(c) Discretionary
WV W. Va. R. Crim. P. 7(f) W. Va. Code Ann. § 62-1B-1 (LexisNexis 2014) Discretionary
WI State ex rel. Drew v. Shaughnessy, 249 N.W. 522, 527 (Wis. 1933) Discretionary
WY Wyo. R. Crim. P. 3(b), (f) Discretionary
FED Fed. R. Crim. P. 7(c)(1), (f) Discretionary

Table E: The Law of Amendment and Dismissal#

# This table supports Figure 18, supra (p. 1364) and reports the law of amendment and dismissal. As noted supra note 186, the difference between amendment and dismissal is slight insofar as prosecutorial charge-sliding is concerned: The former refers to replacing one charge with a lesser one, whereas the latter refers to eliminating the charge altogether. Both maneuvers are charge reductions—dismissal simply reduces the charge to zero. In most instances, the law of dismissal will be the more significant regulatory lever. For one thing, whenever the law of joinder–severance affords the prosecutor sufficient leeway to stack up a large pile of charges, she will be able to generate the most leverage by filing the maximal number of charges possible and then trading them away (that is, dismissing them) piecemeal in exchange for the defendant’s plea. In such a world, the law of amendment will play only an ancillary role. By contrast, if the law of joinder–severance restricts the prosecutor (for example, by authorizing only one charge per case), then the law of amendment may be somewhat more significant, as a prosecutor hoping to slide from one charge to another may need to replace her initial charge with a new one (that is, amend it). But even here, the prosecutor may be able to accomplish the same end by dismissing the first charge and filing a new one that supports the plea deal. In short, as between the law of amendment and the law of dismissal, the latter matters more for purposes of the discussion here. For the sake of simplicity, the substantive codes in this table thus taxonomize the law of dismissal, leaving occasional discussion of the law of amendment to footnotes. Two additional notes of clarification: First, where a state employs an express dismissal rule—that is, a rule authorizing judges to reject charge dismissals in the course of reviewing a proposed plea agreement—the table classifies the state based on that provision, given its direct connection to charge-sliding. Finally, if a state employs either a partial ban on plea bargaining or capped charge discounts in some set of cases, the table employs the codes “partial ban” and “capped,” followed by a parenthetical code describing the dismissal regime that exists in cases not affected by the ban or the cap.

State Source of Authority Procedural Regime
Court Rule Statute Case Law
AL Ala. Code § 15-8-130 (LexisNexis 2015) General Discretion 330 In addition to the cited discretionary dismissal rule, Alabama bars amendment when an “additional or different offense is charged,” Ala. R. Crim. P. 13.5(a), which the state courts interpret strictly to bar amendment whenever the elements of the new offense are not included within the original offense, see Fleming v. State, 814 So. 2d 310, 311 (Ala. Crim. App. 2001).
AK Alaska R. Crim. P. 43(a) Unlimited 331 In addition to the cited unlimited-dismissal rule, Alaska bars amendment when an “additional or different offense is charged,” Alaska R. Crim. P. 7(e), but does not appear to have definitively construed the meaning of the phrase. Cf. McGahan v. State, 606 P.2d 396, 397 n.3 (Alaska 1980) (declining to “express . . . [an] opinion on whether the amendment charged a different offense”).
AZ Ariz. R. Crim. P. 16.4(a) General Discretion
AR Ark. R. Crim. P. 25.3(b) Express Review
CA Cal. Penal Code §§ 999f(b)–(c), 1192.7(a), 1385(a) (2018) Partial Ban
(General Discretion) 332 California bans plea bargaining for specific offenses, including violent sex crimes and driving under the influence, and for certain defendants, including career offenders. Aside from its partial bans, California permits dismissal under general judicial review, Cal. Penal Code § 1385(a), and unlimited amendment before the defendant pleads, id. § 1009.
CO Colo. R. Crim. P. 11(f)(5) Express Review
CT Conn. R. Super. Ct. §§ 36-17, 39-29 Unlimited
DE Del. Super. Ct. R. Crim. P. 48 Unlimited 333 Aside from its unlimited dismissal rule, Delaware permits amendment subject to judicial review unless an “additional or different offense is charged,” Del. Super. Ct. R. Crim. P. 7(e), a phrase the state supreme court appears to interpret so as to permit amendment only to a lesser-included offense of the initial charge, see Commissioner Report & Recommendation at *4, State v. Matos, No. 1003000386, 2015 WL 739954 (Del. Super. Ct. Feb. 19, 2015) (“Under the circumstances of this case, the defendant had adequate notice of the underlying facts of the charge, there were no new elements of the offense added, the amendment was a less serious lesser-included offense of the original charge, and no additional or different offense was being charged.”), aff’d, 124 A.3d 1016 (Del. 2015); State v. Grossberg, No. 9611007818, 1998 WL 278391, at *1 (Del. Super. Ct. Apr. 13, 1998).
FL Barnett v. Antonacci, 122 So. 3d 400, 406 (Fla. Dist. Ct. App. 2013) Unlimited
GA Ga. Code Ann. § 17-8-3 (2013) General Discretion
HI Haw. R. Penal P. 7(f), 48(a) Haw. Rev. Stat. Ann. § 806-56 (LexisNexis 2016) General Discretion
ID Idaho Crim. R. 11(f) Express Review
IL People v. Gill, 886 N.E.2d 1043, 1047 (Ill. App. Ct. 2008) General Discretion
IN Ind. Code §§ 35-34-1-5(c)–(d), -13(a) (2017) Unlimited
IA Iowa Ct. R. 2.33(1) General Discretion 334 See also Iowa Ct. R. 2.10(2) (permitting a court to reject a plea agreement “if the agreement is conditioned upon concurrence of the court in the charging or sentencing concession made”).
KS Kan. Stat. Ann. § 22-3015(b)(2) (2016 Supp.) Express Review
KY Ky. R. Crim. P. 9.64 General Discretion 335 The cited rule affords judges discretion to approve or reject charge dismissals without expressly inviting them to use that authority to review charge bargains; case law in the state does, however, make the invitation explicit. See Hoskins v. Maricle, 150 S.W.3d 1, 24 (Ky. 2004) (“A ‘charge bargain,’ which dismisses or amends one or more charges in exchange for a guilty plea to the reduced charges . . . can be approved or rejected in the discretion of the trial court . . . .”). In addition to the discretionary dismissal rule, the state permits amendment subject to judicial approval so long as “no additional or different offense is charged.” Ky. R. Crim. P. 6.16; cf. Crouch v. Commonwealth, 323 S.W.3d 668, 672 (Ky. 2010) (“[C]hanging the charge . . . from the felony offense of theft of identity to the misdemeanor charge of giving a false name to a peace officer would have resulted in . . . an entirely different offense.”).
LA La. Code Crim. Proc. Ann. art. 691 (2003) Unlimited
ME Me. R. Unified Crim. P. 48(a) Unlimited
MD Md. R. 4-247(a) Unlimited 336 See also Md. R. 4-243(a)(1)(F), (c) (permitting parties to “submit a plea agreement proposing a particular sentence, disposition, or other judicial action to a judge for consideration” and approval).
MA Mass. R. Crim. P. 12(b)(5), (c)(5) Express Review 337 See also Commonwealth v. Gendraw, 774 N.E.2d 167, 174 (Mass. App. Ct. 2002) (“A judge is afforded wide discretion in determining whether to accept a guilty plea.”).
MI Mich. Comp. Laws Ann. § 333.7415 (West 2012) Capped Charge Discounts
(General Discretion) 338 Michigan imposes a limited capped regime for defendants charged with serious drug offenses. See Mich. Comp. Laws Ann. § 333.7415. Under that capped regime, a defendant charged with one of the specified offenses may not have his charges reduced at all prior to arraignment, id. § 333.7415(1), and after arraignment may plead guilty only to a charge from specific proximate tiers of offenses, id. § 333.7415(2). Outside of this capped regime, courts have general discretion to approve or reject charge dismissals. Id. § 767.29.
MN Minn. R. Crim. P. 30.01 Minn. Stat. § 631.21 (2017) General Discretion
MS Miss. R. Crim. P. 14.6(a) Miss. Code Ann. § 99-15-53 (2015) General Discretion
MO Mo. Sup. Ct. R. 24.02(d)–(e) Express Review
MT Mont. Code Ann. § 46-12-211 (2017) Express Review
NE Neb. Rev. Stat. § 29-1606 (2016) General Discretion 339 See also State v. Sanchell, 216 N.W.2d 504, 508 (Neb. 1974) (holding that Neb. Rev. Stat. § 29-1606 “requires court approval [for a plea agreement] to be binding”).
NV Nev. Rev. Stat. §§ 174.085, 212.189, 484C.430, 484C.440, 484C.470 (2015) Partial Ban
(General Discretion) 340 Nevada bans plea bargaining for DUI-related offenses, see Nev. Rev. Stat. §§ 484C.430, 484C.440, 484C.470, and for unlawful acts relating to human excrement or bodily fluid, see id. § 212.189. Aside from its partial bans, Nevada permits dismissal of an indictment or information following the defendant’s arrest only subject to general judicial discretion. See id. § 174.085(5), (7).
NH State ex rel. Bokowsky v. Rudman, 274 A.2d 785, 786 (N.H. 1971) Unlimited
NJ N.J. Ct. R. 3:25-1 General Discretion
NM State v. Sweat, 433 P.2d 229, 230 (N.M. Ct. App. 1967) Unlimited
NY N.Y. Crim. Proc. Law § 220.10(5) (McKinney 2004) Capped Charge Discounts 341 Aside from its capped charge-sliding rule, New York permits dismissal of charges only subject to judicial review. N.Y. Crim. Proc. Law § 210.40(3) (McKinney 2007).
NC N.C. Gen. Stat. § 15A-931 (2017) Unlimited
ND N.D. R. Crim. P. 11(c)(3) Express Review
OH Ohio R. Crim. P. 48(A) Ohio Rev. Code Ann. § 2941.33 (West 2006) General Discretion
OK Okla. Stat. Ann. tit. 22, §§ 815–816 (West 2003) General Discretion
OR Or. Rev. Stat. § 135.755 (2015) General Discretion 342 See also Or. Rev. Stat. § 135.432(2) (“If a tentative plea agreement has been reached which contemplates . . . charge or sentence concessions . . . the trial judge, upon request of the parties, may . . . advise the district attorney and defense counsel whether the trial judge will concur in the proposed disposition . . . .”).
PA Pa. R. Crim. P. 585(A), 586 42 Pa. Stat. and Cons. Stat. Ann. § 8932 (West 2017) General Discretion
RI R.I. Super. Ct. R. Crim. P. 48(a) Unlimited
SC State v. Ridge, 236 S.E.2d 401, 402 (S.C. 1977) Unlimited
SD S.D. Codified Laws §§ 23A-7-8, -9 (2018) Express Review
TN Tenn. R. Crim. P. 11(c)(3) Express Review
TX Tex. Code Crim. Proc. art. 32.02 (2017) General Discretion
UT Utah R. Crim. P. 11(h)(1) Express Review
VT Vt. R. Crim. P. 11(e)(2) Vt. Stat. Ann. tit. 13, § 6565 (2009) Express Review
VA Va. Sup. Ct. R. 3A:8(c) Express Review
WA Wash. Super. Ct. Crim. R. 8.3(a) General Discretion
WV W. Va. R. Crim. P. 11(e)(2) Express Review
WI State v. Kenyon, 270 N.W.2d 160, 164 (Wis. 1978) General Discretion 343 See also Wis. Stat. § 967.055(2) (2018) (expressly encouraging charge-bargain review in drunk-driving cases).
WY Wyo. R. Crim. P. 11(e)(2) Express Review
FED Fed. R. Crim. P. 11(c)(3) Express Review

Table F: State Plea and Prosecution Rates††

†† This table supports Tables 2 and 3, supra (pp. 1375, 1376), and text accompanying supra notes 214–219, by reporting statistics about felony-case dispositions for eighteen states in which such data are readily and reliably available. See supra note 222 and accompanying text (discussing data and limitations). Data for each state roughly correspond to the year 2013, with variation based on local fiscal-year definitions and reporting practices. For sources, see the notes accompanying each state. Percentages reported in parentheses describe each column’s total as a proportion of the total number of felony dispositions. For the guilty-plea column, the number of guilty pleas per 100 crimes is provided parenthetically below the percentage of cases disposed by plea.

State Felony Dispositions Felony
Guilty Pleas
Crimes Reported 344 Crime report data come from the Federal Bureau of Investigation’s Uniform Crime Reports for 2013 and includes all violent crime and property crime. See FBI, State-by-State and National Crime Estimates by Year(s), Unif. Crime Reporting Statistics, [] (last modified Jan. 26, 2017). For the report produced by the data tool, see (last visited Mar. 6, 2018). Each state’s population, as reported by that same source, is provided parenthetically below the total number of reported crimes.
AK 6,284 345 See Alaska Court System, Annual Report FY 2013, at 91 tbl.4.10 (2014), []. 4,738
227 346 Of these, 7 were bench trials (6 convictions, 1 acquittal) and 220 were jury trials (182 convictions, 38 acquittals). Id.
1,289 347 Id.
AZ 47,127 348 Ariz. Courts, Superior Court Case Activity: Fiscal Year 2013, Narrative Summary 5, [] (last visited Mar. 5, 2018). 38,540 349 This figure is not directly reported in the source, but is calculated by subtracting trials and dismissals from total dispositions. See id. at 3, 5. Additionally, the source does report the total number of individuals sentenced (39,563). See id. at 5. From that number one can subtract the difference between the reported trials and the reported acquittals to obtain the number of trial convictions—which can then be subtracted from the number of people sentenced (that is, convicted) in order to yield the number of non-trial convictions (that is, guilty pleas). This calculation also produces a plea tally of 38,540. That is to say, 39,563 reported people sentenced – [(1,147 reported trials) – (124 reported acquittals)] = 38,540 guilty pleas. See id. at 3, 5 (reporting trials and acquittals).


1,147 350 This figure reports the total number of trials for all criminal cases, not just felony cases. However, felony cases account for 95.6% of all reported criminal cases (47,667 out of 49,872). See id. at 4. Of the reported trials, 135 were bench trials and 1,012 were jury trials. Id. at 3. Acquittals accounted for 124 of the total number of trials. Id. at 5.
CA 241,238 351 See Judicial Council of Cal., 2014 Court Statistics Report, Statewide Caseload Trends 2002–2003 Through 2012–2013, at 75 (2014),
2014-Court-Statistics-Report.pdf [].
195,389 352 See id. at 116 tbl.8a.
5,525 353 This figure combines the number of felony cases resolved by jury trial (4,925) and by bench trial (600). See id. at 81.
40,324 354 This figure combines the number of cases disposed “other” than by plea of guilt before a preliminary hearing (28,291) and after a preliminary hearing (12,033). Id. at 116 tbl.8a. The source describes these categories as including “dismissals and transfers.” Id. at 117.
FL 185,986 355 Fla. Office of the State Courts Admin’r, Circuit Criminal Statistics, Florida’s Trial Courts Statistical Reference Guide FY 2012–13, at 3-15 (2014),
fileparse.php/250/urlt/reference-guide-1213-circuit-crim.pdf [].
141,839 356 This figure omits two sizable disposition categories reported by the state: transferred cases (9,808) and “other,” a category containing 15,249 dispositions that combines together “defendants placed on deferred prosecution or other pretrial diversion or probation programs . . . , defendants declared mentally or physically unable to stand trial, cases disposed upon estreature of a bond, cases that are nolle prosequi . . . for administrative reasons, or similar matters.” Id at 3-20. Note that this category includes some cases that approximate guilty pleas (for example, deferred prosecution and diversion agreements) and others that approximate dismissals (for example, cases that end in nolle prosequi for administrative reasons). See id.
3,681 357 Of the total number of trials, 3,361 were to a jury, of which 2,197 ended in a verdict of guilt, 899 ended in either acquittal or dismissal, and 265 ended in a plea. Id. The remaining 320 cases were resolved by bench trial, of which 67 ended in conviction, 50 ended in acquittal or dismissal, and 203 ended in a plea. Id.
15,401 358 Id.
HI 359 See The Judiciary of the State of Hawai’i, 2013 Annual Report: Statistical Supplement 9 tbl.7 (2014),
Statistical_Supp_2013.pdf []. To best approximate felony dispositions, the figures in this row exclude the following categories of offense tabulated in the report: “Driving Under the Influence,” “Liquor Laws,” “Disorderly Conduct,” “Traffic Offenses,” and the miscellaneous category captioned “All Other Offenses (Except Traffic).” See id.
2,753 360 See id. 1,824 361 This figure reports cases terminated with “no trial held,” a category exclusive of trials and dismissals. See id.
151 362 This figure includes 54 non-jury trials and 97 jury trials. See id.
413 363 See id. at 14 tbl.12.
IL 82,464 364 See Admin. Office of the Ill. Courts, Annual Report of the Illinois Courts: Statistical Summary—2013, at 59,
2013/StatsSumm/2013_Statistical_Summary.pdf [] (last visited Mar. 6, 2018).
50,851 365 Id.
3,754 366 This figure combines the reported number of cases in which the defendant was convicted by a jury (658), acquitted by a jury (250), convicted at a bench trial (1,662), and acquitted at a bench trial (1,184). Id.
27,859 367 This figure is reported as the “remaining balance” after trials and guilty pleas are removed, a category that “includes such dispositions as transfers to inactive/fugitive warrant calendar, extradition proceedings, and dismissed on motion of state.” Id.
IN 368 See generally 2 Ind. Judicial Service Report: 2013 Caseload Statistics (2014), [
D7TQ-BBU4]. The figures reported here are based on all felony codes. See id. at 10 (listing codes).
75,245 369 See id. at 170. 57,950 370 This figure combines the number of cases resolved by guilty plea (56,009) and the number of cases resolved by diversion (1,941). See id. at 411, 427; see also id. at 6 (applying a diversion code “[i]f a prosecutor and defendant agree to defer prosecution or for the defendant to enter a diversion program”).
1,573 371 This figure combines the number of cases resolved by jury trial (1,011) and the number of cases resolved by bench trial (562). See id. at 283, 314.
13,465 372 This figure reports dismissals, see id. at 378, which the report defines as “cases which are dismissed either by the court on its own motion (Indiana Trial Rule 41(E)), upon the motion of a party, or upon an agreed entry as the result of settlement between the parties,” id. at 6. It omits an additional 1,310 cases described as “bench dispositions,” see id. at 346, which the report defines as “[c]ases that are disposed by final judicial determination of an issue, but where no witnesses are sworn and no evidence is introduced,” including “hearings on . . . dispositive motions,” id. at 6.
KS 18,239 373 Annual Report of the Courts of Kansas: Fiscal Year 2013, Summary of Felony Caseload for the State 1 (2013),
20Caseload%20Summary.pdf [].
14,618 374 This figure reports the combination of guilty pleas (13,778) and diversionary agreements (840). See id.
511 375 This figure reports the combination of jury trials (414) and bench trials (97). Of all trials, 369 ended in verdicts of guilt, 118 ended in acquittals, 3 ended in pleas after the commencement of trial, and 21 ended in mistrials. See id.
2,624 376 Id.
KY 31,377 377 Admin. Office of the Courts, Report: Statewide Statistics 2, in Annual Report 2013 (2015), [http:
19,534 378 This figure reports the number of felony convictions produced without a trial. Id.
1,260 379 This figure reports the combination of bench trials (1), jury trials (204), and “mixed trial types” (1,055). Id. Of these, the bench trial did not result in conviction, 110 of the jury trials resulted in conviction, and 912 of the mixed trials resulted in conviction. Id.
10,583 380 This figure reports the number of felony cases resolved without a trial (30,117) minus the number of felony convictions produced without a trial (19,534). Id.
MO 381 See Missouri Judicial Report Supplement: Fiscal Year 2013, at 180 tbl.50, 182 tbl.51, [] [hereinafter Missouri Judicial Report Supplement 2013] (last visited Mar. 6, 2018). Table 50 reports cases resolved in the circuit court. Table 51 reports cases resolved in the associate division of the circuit court, where associate judges conduct preliminary hearings. See id. at 320 (providing definitions); see also Circuit Court Judges and Commissioners, Missouri Courts, (on file with the Columbia Law Review) (last visited Mar. 10, 2018) (“In felony criminal cases, the associate circuit judge conducts a preliminary hearing to determine whether there is probable cause to find that a felony has been committed and that the defendant committed it. If probable cause is found, the defendant is ‘bound over’ for trial in the circuit court.”). 49,499 382 This figure reports the total number of felony cases disposed in the circuit courts (44,754) plus the number of cases disposed by plea in the associate division (4,745). See Missouri Judicial Report Supplement 2013, supra note 381, at 181 tbl.50, 183 tbl.51. The figure does not report cases dismissed in the associate division for the reason stated infra note 385. 41,555 383 This figure reports the combination of felony cases disposed by plea in the circuit courts (36,810) and felony cases disposed by plea in the associate division (4,745). See id.
741 384 This figure reports the combination of jury trials (499) and bench trials (242). See id. at 181 tbl.50.
5,322 385 This figure reports the combination of cases dismissed by the prosecutor (5,171) and by the court (151). See id. It does not include an additional 9,338 cases dismissed in the associate division, some of which may have been refiled via indictment. See id. at 183 tbl.51; see also State v. Thomas, 529 S.W.2d 379, 382 (Mo. 1975) (“If the magistrate discharges the accused after the preliminary examination, the prosecuting attorney is without authority to file an information for the offense, but the discharge . . . does not bar a prosecution of the alleged offense by indictment.” (internal quotation marks omitted) (quoting 21 Am. Jur. 2d Criminal Law § 450 (1965))); cf. supra section II.A.3 (describing refiling after preliminary-hearing dismissal).
NM 386 See Admin. Office of the Courts, N.M. Judiciary, Statistical Addendum to the 2013 Annual Report 17, 67,
FY2013_Statistical_Addendum.pdf [] (last visited Mar. 6, 2018) (reporting case dispositions in the general-jurisdiction district courts and the limited-jurisdiction magistrate courts, the latter of which conducts preliminary hearings in felony cases and is empowered to accept guilty pleas at those proceedings).
25,781 387 This figure reports the number of felony cases disposed in the district courts, across the following reported felony categories: crimes against property (7,563), crimes against persons (6,814), drug offenses (5,705), driving while intoxicated (1,367), homicide (139), public safety offenses (842), sexual offenses (634), domestic violence (452), and three categories labeled “habitual offender” (2), “first degree” (55), and “miscellaneous” (852). See id. at 17. The figure also includes the number of felony guilty pleas entered in the magistrate courts (1,356). See id at 67. 19,016 388 This figure reports the combination of felony guilty pleas entered in the district courts for the categories listed supra note 387 (7,675), see id. at 17, along with the felony guilty pleas entered in the magistrate courts (1,356), see id. at 67. It also includes 9,985 felony cases disposed in the district court via “post judgment activity,” on the assumption that these dispositions (accounting for roughly 38% of all felony dispositions) encompass a form of diversion. See id. at 17.
472 389 This figure reports the combination of jury trials (446) and bench trials (26). See id.
6,178 390 This figure reports the combination of cases “dismissed before trial” in the district court (1,968) and cases “dismissed by prosecution” in the district court (4,210). See id. at 17. It omits cases dismissed in the magistrate court at or before a preliminary hearing, as such cases could potentially be refiled in the district court notwithstanding such dismissal. See N.M. Mag. Ct. R. Crim. P. 6-202(D)(1) (“If, upon completion of the examination, the court finds that there is no probable cause to believe that the defendant has committed a felony offense, the court shall dismiss without prejudice . . . .”); cf. supra section II.A.3 (describing the refiling of charges after dismissal at a preliminary hearing).
NY 50,140 391 N.Y. Unified Court Sys., Annual Report 2013, at 24 tbl.5, http://www. [] (last visited Mar. 6, 2018).
43,584 392 Id.
2,045 393 This figure reports a combination of jury convictions (1,214), jury acquittals (460), and nonjury verdicts (371). See id.
3,336 394 Id.
NC 156,991 395 See N.C. Judicial Branch, Statistical Operational Report: Trial Courts, July 1, 2012–June 30, 2013, at 3, 7, [] (last visited Mar. 6, 2018). This figure reports the total number of felony-case dispositions in superior court (123,542), see id. at 3, and the number of felony-case dispositions in district court (33,449) exclusive of preliminary hearing dispositions, see id. at 7. 104,980 396 This figure reports the combination of felony guilty pleas (71,260) and felony deferred prosecution dismissals (776) in the superior courts, see id. at 5, as well as felony guilty pleas (31,055) and felony deferred prosecution dismissals (1,889) in the district courts, see id. at 7.
2,615 397 This figure reports the combination of superior court jury trials (2,110), see id. at 5, and district court trials (505), see id. at 7. Of the superior court trials, 1,486 ended in conviction, 556 ended in acquittal, and 68 ended in guilty pleas before verdict. Id. at 5. Of the district court trials, 446 ended in conviction and 59 ended in acquittal. Id. at 7.
49,396 398 This figure reports the combination of superior court dismissals “with leave” (1,655), dismissals “without leave” (37,900), and dispositions categorized as “other” (9,841). Id. at 5 (noting that the “other” disposition category “includes speedy trial dismissals”). On the distinction between dismissals with and without leave, see N.C. Gen. Stat. §§ 15A-931, -932 (2017).
PA 399 See Kim. E. Nieves et al., Admin. Office of Pa. Courts, 2013 Caseload Statistics of the Unified Judicial System of Pennsylvania 14–15 [hereinafter Pennsylvania 2013 Caseload Statistics], [http://] (last visited Mar. 6, 2018). This source does not distinguish between felonies and misdemeanors when reporting statewide disposition statistics. See id. at 11. Accordingly, the figures in the table pertain exclusively to the Court of Common Pleas in Philadelphia, in which 93.9% of the cases filed are felony cases. See id. at 13. Philadelphia is also home to the Philadelphia Municipal Court, a nonjury forum that is authorized to hear all misdemeanor cases and (with the parties’ consent) all felony cases with maximum potential sentences below five years. See Pa. R. Crim. P. 1001. Unfortunately, disposition data for the Municipal Court similarly fail to distinguish between felony and misdemeanor cases. See Pennsylvania 2013 Caseload Statistics, supra, at 164. Taking both felony and misdemeanor cases together (and excluding cases held for a jury trial in the Court of Common Pleas) the Philadelphia Municipal Court disposed of 39,218 cases, of which 5,812 (14.8%) were resolved by guilty plea; 7,347 (18.7%) were resolved by diversion; 14,378 (36.7%) were withdrawn or dismissed; and 8,967 (22.9%) were resolved by bench trial. Id. As for cases resolved in the Court of Common Pleas outside of Philadelphia, taking felony and misdemeanor cases together, these courts disposed of 164,373 cases, of which 110,955 (67.5%) were resolved by guilty plea; 34,229 (20.8%) were resolved by diversion; 3,549 (2.2%) were resolved by trial; and 8,398 (5.1%) were dismissed. Id. at 11.
15,548 400 Id. at 12. 10,619 401 This figure reports the combination of guilty pleas (10,252) and cases resolved by diversion agreements (367). Id. at 14.
2,231 402 This figure reports the combination of jury trials (657), id. at 15, and nonjury trials (1,574), id. at 14.
74,850 403 This figure pertains to Philadelphia only. See Table 8: Pennsylvania, 2012, FBI: Unif. Crime Reporting,
8tabledatadecpdf/table-8-state-cuts/table-8-pennsylvania [] (last visited Mar. 5, 2018).
(1,538,957) 404 This figure pertains to Philadelphia only. See id.
TN 405 See Annual Report of the Tennessee Judiciary Fiscal Year 2013–2014: Statistics 33, 40, 52, 63, 71, 78, 85, 97, 107, 118, 125, 137, 150, 157, 169, 179, 189, 196, 206, 215, 225, 236, 247, 258, 270, 280, 290, 300, 310, 319, 329,
docs/annual_report__fy2014.pdf [] (last visited Mar. 6, 2018) (reporting disposition data by district). In an attempt to approximate felony case-disposition data, the table reports the following classes of offense: assault, burglary/theft, drugs, homicide, kidnapping, robbery, and sexual offense. The source does not provide direct statewide statistics at the disposition level. Accordingly, the figures presented here are derived from offense-specific disposition data reported for each of the thirty-one districts. A spreadsheet reporting the underlying arithmetic is available at
wp-content/uploads/2018/05/Tennessee-2013-final.xlsx, [hereinafter Tennessee Data Analysis].
76,433 406 Tennessee Data Analysis, supra note 405. 42,601 407 This figure reports the combination of guilty pleas (38,198) and diversion (4,403). Id.
1,381 408 This figure reports the combination of acquittals (358) and convictions after trial (1,023). Id.
30,058 409 Id.
TX 201,013 410 See Office of Court Admin., Annual Statistical Report for the Texas Judiciary: Fiscal Year 2013, at 44 (2014), []. This figure reports the combination of felony guilty pleas (105,946), dismissals (41,037), deferred adjudications (49,257), and trials (4,773); it excludes dispositions on “motions to revoke” and dispositions categorized as “all other dispositions.” Id. 155,203 411 This figure reports the combination of guilty pleas (105,946) and deferred adjudications (49,257). See id.
4,773 412 This figure reports the combination of jury convictions (2,263), jury acquittals (562), bench-trial convictions (1,666), and bench-trial acquittals (282). See id.
41,037 413 See id.
VT 414 See Vt. Judiciary, Annual Statistical Report for FY 2014, at Appendix I, [] (last visited Mar. 7, 2018). The figures in this row report data from fiscal year 2013. 3,370 415 Id. at 1. 2,582 416 Id.
54 417 This figure reports the combination of bench trials (3) and jury trials (51). Id.
WI 34,728 418 Wis. Court Sys., Felony Disposition Summary by Disposing Court Official, Statewide Report (2014), []. 25,969 419 Id.
912 420 This figure reports the combination of jury trials (851) and bench trials (61). Id.
6,835 421 Id. (19.7%) 141,812 (5,742,953)