A defendant’s right to confront the witnesses against him is a cornerstone of our adversarial system of criminal justice. Or is it? Under current law, defendants can invoke their confrontation right only by going to trial. But trials account for about five percent of criminal convictions. That means that the overwhelming majority of defendants convicted in the United States never get to exercise their constitutional right to confront the government’s witnesses.

This Essay argues that the Supreme Court should align its Confrontation Clause jurisprudence with the reality of contemporary criminal justice. The Sixth Amendment grants a criminal defendant the right “to be confronted with the witnesses against him.” The problem is that the Court reads this text as if it said “the witnesses against him at trial.” Nothing compels the Court’s trial-centric gloss on what it means to be a witness. To the contrary, the Confrontation Clause’s text and purposes point towards recognizing that those whose “testimony” the government relies on in plea bargaining are “witnesses” too. This Essay therefore proposes a procedural device through which defendants could exercise their right to confront (i.e., cross-examine) that class of witnesses—the “Sixth Amendment deposition.” By conducting Sixth Amendment depositions, defendants would learn the strengths and weaknesses of the government’s evidence, enabling them to negotiate fairer and more reliable plea bargains. Sixth Amendment depositions would deliver to our “system of pleas” what confrontation at trial brought to an earlier version of American criminal justice—adjudication enhanced by adversarial testing of the government’s case.

The full text of this Essay can be found by clicking the PDF link to the left.


In Crawford v. Washington, the Supreme Court grandly declared that the Confrontation Clause of the Sixth Amendment “commands” that the government’s case against a criminal defendant be tested in the “crucible of cross-examination.” 1 541 U.S. 36, 61 (2004). To enforce that edict, the Court held that prosecutors may not use uncrossed “testimonial hearsay” in criminal trials. 2 Id. at 61–62. This holding is widely viewed as seismic. Crawford “revolutionized this area of the law,” a leading scholar noted recently. 3 Michael S. Pardo, Confrontation After Scalia and Kennedy, 70 Ala. L. Rev. 757, 763 (2019) (“Crawford revolutionized this area of law by effectively severing the ties between the Confrontation Clause’s requirements and modern hearsay law.”); see also Andrew C. Fine, Refining Crawford: The Confrontation Clause After Davis v. Washington and Hammon v. Indiana, 105 Mich. L. Rev. First Impressions 11, 11 (2006) (“[I]n Crawford v. Washington, the Supreme Court had worked a revolutionary transformation of Confrontation Clause analysis . . . .”); Brandon L. Garrett, Constitutional Law and the Law of Evidence, 101 Cornell L. Rev. 57, 69 (2015) (“The Supreme Court’s recent Confrontation Clause jurisprudence over the past few decades has been nothing short of a revolution.”). As another put it, “Crawford is among the most important constitutional cases in modern times.” 4 Gary Lawson, Confronting Crawford: Justice Scalia, the Judicial Method, and the Adjudicative Limits of Originalism, 84 U. Chi. L. Rev. 2265, 2266 (2017).

This Essay argues that the conventional understanding of Crawford as “revolutionary” misses the forest for the trees. Crawford might have transformed Confrontation Clause jurisprudence, but it left the confrontation right itself mostly dormant. That’s because the confrontation right the Court articulated in Crawford is one that defendants can claim only at trial, and our criminal legal system has remarkably few of those. Criminal convictions today come overwhelmingly from guilty pleas, not guilty verdicts. 5 See infra note 10 and accompanying text. In a system of pleas, 6 See Lafler v. Cooper, 566 U.S. 156, 170 (2012) (“[C]riminal justice today is for the most part a system of pleas, not a system of trials.”). a confrontation right that attaches at trial cannot make good on Crawford’s pledge to restore the “crucible of cross-examination.” 7 Crawford, 541 U.S. at 61.

Judicial and scholarly interest in confrontation has been immense since Crawford. 8 Since 2004, the Court has spent hundreds of pages in the United States Reports fleshing out what Crawford means for criminal trials. For a list of the cases, see infra note 52. Scholars have added tens or hundreds of thousands more. See Sopen B. Shah, Guidelines for Guidelines: Implications of the Confrontation Clause’s Revival for Federal Sentencing, 48 J. Marshall L. Rev. 1039, 1050 (2015) (observing that “Crawford’s revival of the Confrontation Clause inspired an industry’s worth of scholarship”); see also Andrew King-Ries, State v. Mizenko: The Montana Supreme Court Wades into the Post-Crawford Waters, 67 Mont. L. Rev. 275, 280 (2006) (noting two years after Crawford that “[n]early three hundred articles have been written that address the . . . decision”). Yet it has slipped under the radar that plea bargaining renders Crawford and the Confrontation Clause empty promises. 9 Shaakirrah Sanders and Sopen Shah come the closest to exposing the problem. They (separately) argue that the practice of plea bargaining cuts in favor of extending the confrontation right past the trial stage of the criminal process into sentencing hearings. See Shaakirrah R. Sanders, Unbranding Confrontation as Only a Trial Right, 65 Hastings L.J. 1257, 1259 (2014) (“Testing the veracity of testimonial statements that are material to punishment is as compelling at felony sentencing as at trial . . . .”); Shah, supra note 8, at 1063 (“The confrontation right—or the probation department’s mere anticipation of the confrontation requirement when preparing the [pre-sentence report]—could ameliorate . . . situation[s where limited criminal history information was obtained prior to a guilty plea].” (citing Stanley A. Weigel, The Sentencing Reform Act of 1984: A Practical Appraisal, 36 UCLA L. Rev. 83, 94 (1988))). I’m sympathetic to their arguments, but the claims I develop here—that the confrontation right should apply prior to the trial phase, during plea bargaining, and that it entitles defendants to take depositions in aid of the plea bargaining process—are very different. It doesn’t have to be this way. Plea bargaining and a robust confrontation right could coexist. But for the Confrontation Clause to be more than a parchment barrier, the Supreme Court would have to update its doctrine to account for the world as it is—a world in which plea bargaining is ubiquitous. Fortunately, there is precedent for such a move, and it comes from one of the Confrontation Clause’s closest neighbors. In Lafler v. Cooper and Missouri v. Frye, the Supreme Court updated the Sixth Amendment’s Counsel Clause to make it relevant to today’s criminal legal system. 10 See Lafler, 566 U.S. at 168–70 (acknowledging the ubiquity of pleas in the criminal justice system and extending the right to the effective assistance of counsel to plea negotiations); Missouri v. Frye, 566 U.S. 134, 143 (2012) (“The reality is that plea bargains have become so central to the administration of the criminal justice system that defense counsel have responsibilities in the plea bargain process, responsibilities that must be met to render the adequate assistance of counsel that the Sixth Amendment requires . . . .”). This Essay shows how the Court could do the same for the Confrontation Clause.

Here’s the core argument. The Confrontation Clause provides that “[i]n all criminal prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses against him.” 11 U.S. Const. amend. VI. Crawford recognized that the crucial question in implementing this text is figuring out who counts as a “witness against” an accused. 12 Crawford, 541 U.S. at 42–43. The Court answered the question by defining a witness for purposes of the Confrontation Clause as  one  who  (1)  makes  a  “testimonial”  statement  that  (2)  the  government  uses  at trial. 13 See infra section II.B. I take the “testimonial” component of the definition—which has generated a lion’s share of the scholarly attention and case law—as given. Instead, I challenge the requirement that the statement be used at trial. 14 See infra section III.C.

The Court’s trial-centric gloss on what it means to be a “witness against” an accused is at odds with the reality of the American criminal legal system—a “post-trial world” where the critical adjudicator is usually a prosecutor, not a judge or jury. 15 See infra notes 66–68, 109–113 and accompanying text. See generally Jocelyn Simonson, The Criminal Court Audience in a Post-Trial World, 127 Harv. L. Rev. 2173 (2014) (coining the phrase “post-trial world”). If only “trial witnesses” count as “witnesses,” then in a world without trials there can be few “witnesses” and very little confrontation. The Court’s restrictive gloss on “witnesses” is also inconsistent with the Sixth Amendment’s text, which attaches the confrontation right to those facing “criminal prosecution,” not just those in trial. 16 See infra notes 188–190 and accompanying text. And it departs from the Court’s treatment of the Sixth Amendment in Lafler and Frye, where the Court held that a defendant may assert an ineffective assistance of counsel claim when their lawyer’s deficient performance caused them to miss out on a favorable plea deal. 17 See Lafler v. Cooper, 566 U.S. 156, 169–70 (2012) (rejecting the contention that “[a] fair trial wipes clean any deficient performance by defense counsel during plea bargaining”); Missouri v. Frye, 566 U.S. 134, 145 (2012) (holding that, under the Sixth Amendment, defense counsel must communicate plea offers to their clients). In those cases, unlike in Crawford, the Court rejected the theory that the Sixth Amendment is meant only to ensure a fair trial. 18 See infra section I.B. That was because, as the Court explained in Frye, “ours ‘is for the most part a system of pleas, not a system of trials.’” 19 Frye, 566 U.S. at 143–44 (quoting Lafler, 566 U.S. at 170).

So then who are the “witnesses against” defendants in a system of pleas? They are the people whose “testimony” is used in the only adjudication that routinely matters—the plea bargain. The Confrontation Clause, I argue, commands that the government confront the defendant with those people, and the natural procedural mechanism for such confrontation is a deposition. I therefore propose that in the age of plea bargaining, the Confrontation Clause entitles criminal defendants to take the deposition  of any  accusatory  “witness”  whom  prosecutors  rely  on  in  plea bargaining. 20 See infra section III.A. Because plea bargaining is a fluid process, I look to a fixed point—the filing of charges—to determine the class of people on whom the prosecution has relied. See infra note 209 and accompanying text. Depositions are not a routine part of criminal practice today in the federal system or in most states. 21 See Ion Meyn, Discovery and Darkness: The Information Deficit in Criminal Disputes, 79 Brook. L. Rev. 1091, 1094 (2014) (highlighting that typically, criminal discovery statutes do not grant defendants formal pretrial information-gathering mechanisms such as depositions, interrogatories, and document requests). But a few states already allow criminal defendants to take them, and nothing about the logic of criminal litigation is inconsistent with depositions. 22 See infra notes 283–297 and accompanying text.

The objective for “Sixth Amendment depositions” would be to incorporate adversarial testing into plea bargaining outcomes. If the government’s witnesses performed well under cross-examination, the prosecutor’s negotiating leverage would increase, and with it the price of a plea. If the witnesses did poorly, that would strengthen the defendant’s hand. But whether depositions led to a higher or a lower plea price in any particular case, adversarial testing of the government’s evidence would have contributed to a fairer and more reliable adjudication. 23 See infra notes 247–250 and accompanying text. That is the Confrontation Clause’s purpose. 24 See infra notes 139, 200–201 and accompanying text. By modernizing the confrontation right, by bringing it into the twenty-first century (or just the twentieth), 25 See generally William Ortman, When Plea Bargaining Became Normal, 100 B.U. L. Rev. 1435 (2020) (exploring the intellectual history of plea bargaining in the early- to mid-twentieth century). the Court could restore some of the adversarial process that plea bargaining has upended. 26 See Stephanos Bibas, Incompetent Plea Bargaining and Extrajudicial Reforms, 126 Harv. L. Rev. 150, 164 (2012) (“Plea bargaining today is fundamentally not adversarial but collaborative (some would say collusive).”); see also Gerard E. Lynch, Our Administrative System of Criminal Justice, 66 Fordham L. Rev. 2117, 2121 (1998) [hereinafter Lynch, Our Administrative System] (noting that the pervasiveness of plea bargaining “has resulted in the development of a system of justice that actually looks . . . far more like . . . an inquisitorial system than like the idealized model of adversary justice described in the textbooks”); James E. Pfander & Daniel D. Birk, Article III Judicial Power, the Adverse-Party Requirement, and Non-Contentious Jurisdiction, 124 Yale L.J. 1346, 1472–74 (2015) (acknowledging the historical pedigree of certain nonadversarial functions of the court).

The full argument develops in three Parts. Part I describes the problem in more detail, Part II diagnoses its doctrinal roots, and Part III offers Sixth Amendment depositions as a solution. Part III also considers several likely objections to the proposal—that depositions would be too costly, that they would burden witnesses, and that they would become just one more bargaining chip for prosecutors and defendants to haggle over in plea bargaining—and argues that none justifies preserving the status quo. Finally, a brief Conclusion situates the proposal as an installment payment on the project of constitutional criminal procedure modernization that the Court started in Lafler and Frye.